Home | History | Annotate | Line # | Download | only in libgomp
libgomp.info revision 1.11
      1  1.11  mrg This is libgomp.info, produced by makeinfo version 6.5 from
      2  1.11  mrg libgomp.texi.
      3   1.1  mrg 
      4  1.11  mrg Copyright (C) 2006-2018 Free Software Foundation, Inc.
      5   1.1  mrg 
      6   1.1  mrg    Permission is granted to copy, distribute and/or modify this document
      7   1.3  mrg under the terms of the GNU Free Documentation License, Version 1.3 or
      8   1.1  mrg any later version published by the Free Software Foundation; with the
      9   1.1  mrg Invariant Sections being "Funding Free Software", the Front-Cover texts
     10   1.1  mrg being (a) (see below), and with the Back-Cover Texts being (b) (see
     11   1.1  mrg below).  A copy of the license is included in the section entitled "GNU
     12   1.1  mrg Free Documentation License".
     13   1.1  mrg 
     14   1.1  mrg    (a) The FSF's Front-Cover Text is:
     15   1.1  mrg 
     16   1.1  mrg    A GNU Manual
     17   1.1  mrg 
     18   1.1  mrg    (b) The FSF's Back-Cover Text is:
     19   1.1  mrg 
     20   1.1  mrg    You have freedom to copy and modify this GNU Manual, like GNU
     21  1.11  mrg software.  Copies published by the Free Software Foundation raise funds
     22  1.11  mrg for GNU development.
     23   1.1  mrg INFO-DIR-SECTION GNU Libraries
     24   1.1  mrg START-INFO-DIR-ENTRY
     25   1.5  mrg * libgomp: (libgomp).          GNU Offloading and Multi Processing Runtime Library.
     26   1.1  mrg END-INFO-DIR-ENTRY
     27   1.1  mrg 
     28   1.5  mrg    This manual documents libgomp, the GNU Offloading and Multi
     29   1.5  mrg Processing Runtime library.  This is the GNU implementation of the
     30   1.5  mrg OpenMP and OpenACC APIs for parallel and accelerator programming in
     31   1.5  mrg C/C++ and Fortran.
     32   1.1  mrg 
     33   1.1  mrg    Published by the Free Software Foundation 51 Franklin Street, Fifth
     34   1.1  mrg Floor Boston, MA 02110-1301 USA
     35   1.1  mrg 
     36  1.11  mrg    Copyright (C) 2006-2018 Free Software Foundation, Inc.
     37   1.1  mrg 
     38   1.1  mrg    Permission is granted to copy, distribute and/or modify this document
     39   1.3  mrg under the terms of the GNU Free Documentation License, Version 1.3 or
     40   1.1  mrg any later version published by the Free Software Foundation; with the
     41   1.1  mrg Invariant Sections being "Funding Free Software", the Front-Cover texts
     42   1.1  mrg being (a) (see below), and with the Back-Cover Texts being (b) (see
     43   1.1  mrg below).  A copy of the license is included in the section entitled "GNU
     44   1.1  mrg Free Documentation License".
     45   1.1  mrg 
     46   1.1  mrg    (a) The FSF's Front-Cover Text is:
     47   1.1  mrg 
     48   1.1  mrg    A GNU Manual
     49   1.1  mrg 
     50   1.1  mrg    (b) The FSF's Back-Cover Text is:
     51   1.1  mrg 
     52   1.1  mrg    You have freedom to copy and modify this GNU Manual, like GNU
     53  1.11  mrg software.  Copies published by the Free Software Foundation raise funds
     54  1.11  mrg for GNU development.
     55   1.1  mrg 
     56   1.1  mrg 
     57   1.1  mrg File: libgomp.info,  Node: Top,  Next: Enabling OpenMP,  Up: (dir)
     58   1.1  mrg 
     59   1.1  mrg Introduction
     60   1.1  mrg ************
     61   1.1  mrg 
     62  1.11  mrg This manual documents the usage of libgomp, the GNU Offloading and Multi
     63  1.11  mrg Processing Runtime Library.  This includes the GNU implementation of the
     64  1.11  mrg OpenMP (http://www.openmp.org) Application Programming Interface (API)
     65  1.11  mrg for multi-platform shared-memory parallel programming in C/C++ and
     66  1.11  mrg Fortran, and the GNU implementation of the OpenACC
     67  1.11  mrg (https://www.openacc.org) Application Programming Interface (API) for
     68   1.5  mrg offloading of code to accelerator devices in C/C++ and Fortran.
     69   1.5  mrg 
     70   1.5  mrg    Originally, libgomp implemented the GNU OpenMP Runtime Library.
     71   1.5  mrg Based on this, support for OpenACC and offloading (both OpenACC and
     72   1.5  mrg OpenMP 4's target construct) has been added later on, and the library's
     73   1.5  mrg name changed to GNU Offloading and Multi Processing Runtime Library.
     74   1.1  mrg 
     75   1.1  mrg * Menu:
     76   1.1  mrg 
     77   1.1  mrg * Enabling OpenMP::            How to enable OpenMP for your applications.
     78   1.1  mrg * Runtime Library Routines::   The OpenMP runtime application programming
     79   1.1  mrg                                interface.
     80   1.1  mrg * Environment Variables::      Influencing runtime behavior with environment
     81   1.1  mrg                                variables.
     82   1.7  mrg * Enabling OpenACC::           How to enable OpenACC for your
     83   1.7  mrg                                applications.
     84   1.7  mrg * OpenACC Runtime Library Routines:: The OpenACC runtime application
     85   1.7  mrg                                programming interface.
     86   1.7  mrg * OpenACC Environment Variables:: Influencing OpenACC runtime behavior with
     87   1.7  mrg                                environment variables.
     88   1.7  mrg * CUDA Streams Usage::         Notes on the implementation of
     89   1.7  mrg                                asynchronous operations.
     90   1.7  mrg * OpenACC Library Interoperability:: OpenACC library interoperability with the
     91   1.7  mrg                                NVIDIA CUBLAS library.
     92   1.1  mrg * The libgomp ABI::            Notes on the external ABI presented by libgomp.
     93   1.5  mrg * Reporting Bugs::             How to report bugs in the GNU Offloading and
     94   1.5  mrg                                Multi Processing Runtime Library.
     95   1.1  mrg * Copying::                    GNU general public license says
     96   1.1  mrg                                how you can copy and share libgomp.
     97   1.1  mrg * GNU Free Documentation License::
     98   1.1  mrg                                How you can copy and share this manual.
     99   1.1  mrg * Funding::                    How to help assure continued work for free
    100   1.1  mrg                                software.
    101   1.3  mrg * Library Index::              Index of this documentation.
    102   1.1  mrg 
    103   1.1  mrg 
    104   1.1  mrg File: libgomp.info,  Node: Enabling OpenMP,  Next: Runtime Library Routines,  Prev: Top,  Up: Top
    105   1.1  mrg 
    106   1.1  mrg 1 Enabling OpenMP
    107   1.1  mrg *****************
    108   1.1  mrg 
    109   1.1  mrg To activate the OpenMP extensions for C/C++ and Fortran, the
    110  1.11  mrg compile-time flag '-fopenmp' must be specified.  This enables the OpenMP
    111  1.11  mrg directive '#pragma omp' in C/C++ and '!$omp' directives in free form,
    112  1.11  mrg 'c$omp', '*$omp' and '!$omp' directives in fixed form, '!$' conditional
    113  1.11  mrg compilation sentinels in free form and 'c$', '*$' and '!$' sentinels in
    114  1.11  mrg fixed form, for Fortran.  The flag also arranges for automatic linking
    115  1.11  mrg of the OpenMP runtime library (*note Runtime Library Routines::).
    116  1.11  mrg 
    117  1.11  mrg    A complete description of all OpenMP directives accepted may be found
    118  1.11  mrg in the OpenMP Application Program Interface (http://www.openmp.org)
    119  1.11  mrg manual, version 4.5.
    120   1.1  mrg 
    121   1.1  mrg 
    122   1.1  mrg File: libgomp.info,  Node: Runtime Library Routines,  Next: Environment Variables,  Prev: Enabling OpenMP,  Up: Top
    123   1.1  mrg 
    124   1.1  mrg 2 Runtime Library Routines
    125   1.1  mrg **************************
    126   1.1  mrg 
    127   1.5  mrg The runtime routines described here are defined by Section 3 of the
    128   1.7  mrg OpenMP specification in version 4.5.  The routines are structured in
    129   1.1  mrg following three parts:
    130   1.1  mrg 
    131   1.5  mrg * Menu:
    132   1.1  mrg 
    133   1.5  mrg Control threads, processors and the parallel environment.  They have C
    134   1.5  mrg linkage, and do not throw exceptions.
    135   1.1  mrg 
    136   1.1  mrg * omp_get_active_level::        Number of active parallel regions
    137   1.1  mrg * omp_get_ancestor_thread_num:: Ancestor thread ID
    138   1.5  mrg * omp_get_cancellation::        Whether cancellation support is enabled
    139   1.5  mrg * omp_get_default_device::      Get the default device for target regions
    140   1.1  mrg * omp_get_dynamic::             Dynamic teams setting
    141   1.1  mrg * omp_get_level::               Number of parallel regions
    142   1.3  mrg * omp_get_max_active_levels::   Maximum number of active regions
    143   1.7  mrg * omp_get_max_task_priority::   Maximum task priority value that can be set
    144   1.3  mrg * omp_get_max_threads::         Maximum number of threads of parallel region
    145   1.1  mrg * omp_get_nested::              Nested parallel regions
    146   1.5  mrg * omp_get_num_devices::         Number of target devices
    147   1.1  mrg * omp_get_num_procs::           Number of processors online
    148   1.5  mrg * omp_get_num_teams::           Number of teams
    149   1.1  mrg * omp_get_num_threads::         Size of the active team
    150   1.5  mrg * omp_get_proc_bind::           Whether theads may be moved between CPUs
    151   1.1  mrg * omp_get_schedule::            Obtain the runtime scheduling method
    152   1.5  mrg * omp_get_team_num::            Get team number
    153   1.1  mrg * omp_get_team_size::           Number of threads in a team
    154   1.3  mrg * omp_get_thread_limit::        Maximum number of threads
    155   1.1  mrg * omp_get_thread_num::          Current thread ID
    156   1.1  mrg * omp_in_parallel::             Whether a parallel region is active
    157   1.3  mrg * omp_in_final::                Whether in final or included task region
    158   1.5  mrg * omp_is_initial_device::       Whether executing on the host device
    159   1.5  mrg * omp_set_default_device::      Set the default device for target regions
    160   1.1  mrg * omp_set_dynamic::             Enable/disable dynamic teams
    161   1.1  mrg * omp_set_max_active_levels::   Limits the number of active parallel regions
    162   1.1  mrg * omp_set_nested::              Enable/disable nested parallel regions
    163   1.1  mrg * omp_set_num_threads::         Set upper team size limit
    164   1.1  mrg * omp_set_schedule::            Set the runtime scheduling method
    165   1.1  mrg 
    166   1.5  mrg Initialize, set, test, unset and destroy simple and nested locks.
    167   1.1  mrg 
    168   1.1  mrg * omp_init_lock::            Initialize simple lock
    169   1.1  mrg * omp_set_lock::             Wait for and set simple lock
    170   1.1  mrg * omp_test_lock::            Test and set simple lock if available
    171   1.1  mrg * omp_unset_lock::           Unset simple lock
    172   1.1  mrg * omp_destroy_lock::         Destroy simple lock
    173   1.1  mrg * omp_init_nest_lock::       Initialize nested lock
    174   1.1  mrg * omp_set_nest_lock::        Wait for and set simple lock
    175   1.1  mrg * omp_test_nest_lock::       Test and set nested lock if available
    176   1.1  mrg * omp_unset_nest_lock::      Unset nested lock
    177   1.1  mrg * omp_destroy_nest_lock::    Destroy nested lock
    178   1.1  mrg 
    179   1.5  mrg Portable, thread-based, wall clock timer.
    180   1.1  mrg 
    181   1.1  mrg * omp_get_wtick::            Get timer precision.
    182   1.1  mrg * omp_get_wtime::            Elapsed wall clock time.
    183   1.1  mrg 
    184   1.1  mrg 
    185   1.1  mrg File: libgomp.info,  Node: omp_get_active_level,  Next: omp_get_ancestor_thread_num,  Up: Runtime Library Routines
    186   1.1  mrg 
    187  1.11  mrg 2.1 'omp_get_active_level' - Number of parallel regions
    188   1.1  mrg =======================================================
    189   1.1  mrg 
    190   1.1  mrg _Description_:
    191   1.1  mrg      This function returns the nesting level for the active parallel
    192   1.1  mrg      blocks, which enclose the calling call.
    193   1.1  mrg 
    194   1.1  mrg _C/C++_
    195  1.11  mrg      _Prototype_:   'int omp_get_active_level(void);'
    196   1.1  mrg 
    197   1.1  mrg _Fortran_:
    198  1.11  mrg      _Interface_:   'integer function omp_get_active_level()'
    199   1.1  mrg 
    200   1.1  mrg _See also_:
    201   1.1  mrg      *note omp_get_level::, *note omp_get_max_active_levels::, *note
    202   1.1  mrg      omp_set_max_active_levels::
    203   1.1  mrg 
    204   1.1  mrg _Reference_:
    205   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.20.
    206   1.1  mrg 
    207   1.1  mrg 
    208   1.5  mrg File: libgomp.info,  Node: omp_get_ancestor_thread_num,  Next: omp_get_cancellation,  Prev: omp_get_active_level,  Up: Runtime Library Routines
    209   1.1  mrg 
    210  1.11  mrg 2.2 'omp_get_ancestor_thread_num' - Ancestor thread ID
    211   1.1  mrg ======================================================
    212   1.1  mrg 
    213   1.1  mrg _Description_:
    214   1.1  mrg      This function returns the thread identification number for the
    215   1.5  mrg      given nesting level of the current thread.  For values of LEVEL
    216  1.11  mrg      outside zero to 'omp_get_level' -1 is returned; if LEVEL is
    217  1.11  mrg      'omp_get_level' the result is identical to 'omp_get_thread_num'.
    218   1.1  mrg 
    219   1.1  mrg _C/C++_
    220  1.11  mrg      _Prototype_:   'int omp_get_ancestor_thread_num(int level);'
    221   1.1  mrg 
    222   1.1  mrg _Fortran_:
    223  1.11  mrg      _Interface_:   'integer function omp_get_ancestor_thread_num(level)'
    224  1.11  mrg                     'integer level'
    225   1.1  mrg 
    226   1.1  mrg _See also_:
    227   1.1  mrg      *note omp_get_level::, *note omp_get_thread_num::, *note
    228   1.1  mrg      omp_get_team_size::
    229   1.1  mrg 
    230   1.1  mrg _Reference_:
    231   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.18.
    232   1.1  mrg 
    233   1.1  mrg 
    234   1.5  mrg File: libgomp.info,  Node: omp_get_cancellation,  Next: omp_get_default_device,  Prev: omp_get_ancestor_thread_num,  Up: Runtime Library Routines
    235   1.5  mrg 
    236  1.11  mrg 2.3 'omp_get_cancellation' - Whether cancellation support is enabled
    237   1.5  mrg ====================================================================
    238   1.5  mrg 
    239   1.5  mrg _Description_:
    240  1.11  mrg      This function returns 'true' if cancellation is activated, 'false'
    241  1.11  mrg      otherwise.  Here, 'true' and 'false' represent their
    242  1.11  mrg      language-specific counterparts.  Unless 'OMP_CANCELLATION' is set
    243   1.5  mrg      true, cancellations are deactivated.
    244   1.5  mrg 
    245   1.5  mrg _C/C++_:
    246  1.11  mrg      _Prototype_:   'int omp_get_cancellation(void);'
    247   1.5  mrg 
    248   1.5  mrg _Fortran_:
    249  1.11  mrg      _Interface_:   'logical function omp_get_cancellation()'
    250   1.5  mrg 
    251   1.5  mrg _See also_:
    252   1.5  mrg      *note OMP_CANCELLATION::
    253   1.5  mrg 
    254   1.5  mrg _Reference_:
    255   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.9.
    256   1.5  mrg 
    257   1.5  mrg 
    258   1.5  mrg File: libgomp.info,  Node: omp_get_default_device,  Next: omp_get_dynamic,  Prev: omp_get_cancellation,  Up: Runtime Library Routines
    259   1.5  mrg 
    260  1.11  mrg 2.4 'omp_get_default_device' - Get the default device for target regions
    261   1.5  mrg ========================================================================
    262   1.5  mrg 
    263   1.5  mrg _Description_:
    264   1.5  mrg      Get the default device for target regions without device clause.
    265   1.5  mrg 
    266   1.5  mrg _C/C++_:
    267  1.11  mrg      _Prototype_:   'int omp_get_default_device(void);'
    268   1.1  mrg 
    269   1.5  mrg _Fortran_:
    270  1.11  mrg      _Interface_:   'integer function omp_get_default_device()'
    271   1.5  mrg 
    272   1.5  mrg _See also_:
    273   1.5  mrg      *note OMP_DEFAULT_DEVICE::, *note omp_set_default_device::
    274   1.5  mrg 
    275   1.5  mrg _Reference_:
    276   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.30.
    277   1.5  mrg 
    278   1.5  mrg 
    279   1.5  mrg File: libgomp.info,  Node: omp_get_dynamic,  Next: omp_get_level,  Prev: omp_get_default_device,  Up: Runtime Library Routines
    280   1.5  mrg 
    281  1.11  mrg 2.5 'omp_get_dynamic' - Dynamic teams setting
    282   1.1  mrg =============================================
    283   1.1  mrg 
    284   1.1  mrg _Description_:
    285  1.11  mrg      This function returns 'true' if enabled, 'false' otherwise.  Here,
    286  1.11  mrg      'true' and 'false' represent their language-specific counterparts.
    287   1.1  mrg 
    288   1.1  mrg      The dynamic team setting may be initialized at startup by the
    289  1.11  mrg      'OMP_DYNAMIC' environment variable or at runtime using
    290  1.11  mrg      'omp_set_dynamic'.  If undefined, dynamic adjustment is disabled by
    291  1.11  mrg      default.
    292   1.1  mrg 
    293   1.1  mrg _C/C++_:
    294  1.11  mrg      _Prototype_:   'int omp_get_dynamic(void);'
    295   1.1  mrg 
    296   1.1  mrg _Fortran_:
    297  1.11  mrg      _Interface_:   'logical function omp_get_dynamic()'
    298   1.1  mrg 
    299   1.1  mrg _See also_:
    300   1.1  mrg      *note omp_set_dynamic::, *note OMP_DYNAMIC::
    301   1.1  mrg 
    302   1.1  mrg _Reference_:
    303   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.8.
    304   1.1  mrg 
    305   1.1  mrg 
    306   1.1  mrg File: libgomp.info,  Node: omp_get_level,  Next: omp_get_max_active_levels,  Prev: omp_get_dynamic,  Up: Runtime Library Routines
    307   1.1  mrg 
    308  1.11  mrg 2.6 'omp_get_level' - Obtain the current nesting level
    309   1.1  mrg ======================================================
    310   1.1  mrg 
    311   1.1  mrg _Description_:
    312   1.1  mrg      This function returns the nesting level for the parallel blocks,
    313   1.1  mrg      which enclose the calling call.
    314   1.1  mrg 
    315   1.1  mrg _C/C++_
    316  1.11  mrg      _Prototype_:   'int omp_get_level(void);'
    317   1.1  mrg 
    318   1.1  mrg _Fortran_:
    319  1.11  mrg      _Interface_:   'integer function omp_level()'
    320   1.1  mrg 
    321   1.1  mrg _See also_:
    322   1.1  mrg      *note omp_get_active_level::
    323   1.1  mrg 
    324   1.1  mrg _Reference_:
    325   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.17.
    326   1.1  mrg 
    327   1.1  mrg 
    328   1.7  mrg File: libgomp.info,  Node: omp_get_max_active_levels,  Next: omp_get_max_task_priority,  Prev: omp_get_level,  Up: Runtime Library Routines
    329   1.1  mrg 
    330  1.11  mrg 2.7 'omp_get_max_active_levels' - Maximum number of active regions
    331   1.1  mrg ==================================================================
    332   1.1  mrg 
    333   1.1  mrg _Description_:
    334   1.3  mrg      This function obtains the maximum allowed number of nested, active
    335   1.3  mrg      parallel regions.
    336   1.1  mrg 
    337   1.1  mrg _C/C++_
    338  1.11  mrg      _Prototype_:   'int omp_get_max_active_levels(void);'
    339   1.1  mrg 
    340   1.1  mrg _Fortran_:
    341  1.11  mrg      _Interface_:   'integer function omp_get_max_active_levels()'
    342   1.1  mrg 
    343   1.1  mrg _See also_:
    344   1.1  mrg      *note omp_set_max_active_levels::, *note omp_get_active_level::
    345   1.1  mrg 
    346   1.1  mrg _Reference_:
    347   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.16.
    348   1.1  mrg 
    349   1.1  mrg 
    350   1.7  mrg File: libgomp.info,  Node: omp_get_max_task_priority,  Next: omp_get_max_threads,  Prev: omp_get_max_active_levels,  Up: Runtime Library Routines
    351   1.7  mrg 
    352  1.11  mrg 2.8 'omp_get_max_task_priority' - Maximum priority value
    353   1.7  mrg ========================================================
    354   1.7  mrg 
    355   1.7  mrg that can be set for tasks.
    356   1.7  mrg _Description_:
    357   1.7  mrg      This function obtains the maximum allowed priority number for
    358   1.7  mrg      tasks.
    359   1.7  mrg 
    360   1.7  mrg _C/C++_
    361  1.11  mrg      _Prototype_:   'int omp_get_max_task_priority(void);'
    362   1.1  mrg 
    363   1.7  mrg _Fortran_:
    364  1.11  mrg      _Interface_:   'integer function omp_get_max_task_priority()'
    365   1.7  mrg 
    366   1.7  mrg _Reference_:
    367   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.29.
    368   1.7  mrg 
    369   1.7  mrg 
    370   1.7  mrg File: libgomp.info,  Node: omp_get_max_threads,  Next: omp_get_nested,  Prev: omp_get_max_task_priority,  Up: Runtime Library Routines
    371   1.7  mrg 
    372  1.11  mrg 2.9 'omp_get_max_threads' - Maximum number of threads of parallel region
    373   1.1  mrg ========================================================================
    374   1.1  mrg 
    375   1.1  mrg _Description_:
    376   1.3  mrg      Return the maximum number of threads used for the current parallel
    377  1.11  mrg      region that does not use the clause 'num_threads'.
    378   1.1  mrg 
    379   1.1  mrg _C/C++_:
    380  1.11  mrg      _Prototype_:   'int omp_get_max_threads(void);'
    381   1.1  mrg 
    382   1.1  mrg _Fortran_:
    383  1.11  mrg      _Interface_:   'integer function omp_get_max_threads()'
    384   1.1  mrg 
    385   1.1  mrg _See also_:
    386   1.1  mrg      *note omp_set_num_threads::, *note omp_set_dynamic::, *note
    387   1.1  mrg      omp_get_thread_limit::
    388   1.1  mrg 
    389   1.1  mrg _Reference_:
    390   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.3.
    391   1.1  mrg 
    392   1.1  mrg 
    393   1.5  mrg File: libgomp.info,  Node: omp_get_nested,  Next: omp_get_num_devices,  Prev: omp_get_max_threads,  Up: Runtime Library Routines
    394   1.1  mrg 
    395  1.11  mrg 2.10 'omp_get_nested' - Nested parallel regions
    396   1.7  mrg ===============================================
    397   1.1  mrg 
    398   1.1  mrg _Description_:
    399  1.11  mrg      This function returns 'true' if nested parallel regions are
    400  1.11  mrg      enabled, 'false' otherwise.  Here, 'true' and 'false' represent
    401   1.1  mrg      their language-specific counterparts.
    402   1.1  mrg 
    403   1.1  mrg      Nested parallel regions may be initialized at startup by the
    404  1.11  mrg      'OMP_NESTED' environment variable or at runtime using
    405  1.11  mrg      'omp_set_nested'.  If undefined, nested parallel regions are
    406   1.1  mrg      disabled by default.
    407   1.1  mrg 
    408   1.1  mrg _C/C++_:
    409  1.11  mrg      _Prototype_:   'int omp_get_nested(void);'
    410   1.1  mrg 
    411   1.1  mrg _Fortran_:
    412  1.11  mrg      _Interface_:   'logical function omp_get_nested()'
    413   1.1  mrg 
    414   1.1  mrg _See also_:
    415   1.1  mrg      *note omp_set_nested::, *note OMP_NESTED::
    416   1.1  mrg 
    417   1.1  mrg _Reference_:
    418   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.11.
    419   1.1  mrg 
    420   1.1  mrg 
    421   1.5  mrg File: libgomp.info,  Node: omp_get_num_devices,  Next: omp_get_num_procs,  Prev: omp_get_nested,  Up: Runtime Library Routines
    422   1.1  mrg 
    423  1.11  mrg 2.11 'omp_get_num_devices' - Number of target devices
    424   1.1  mrg =====================================================
    425   1.1  mrg 
    426   1.1  mrg _Description_:
    427   1.5  mrg      Returns the number of target devices.
    428   1.5  mrg 
    429   1.5  mrg _C/C++_:
    430  1.11  mrg      _Prototype_:   'int omp_get_num_devices(void);'
    431   1.5  mrg 
    432   1.5  mrg _Fortran_:
    433  1.11  mrg      _Interface_:   'integer function omp_get_num_devices()'
    434   1.5  mrg 
    435   1.5  mrg _Reference_:
    436   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.31.
    437   1.5  mrg 
    438   1.5  mrg 
    439   1.5  mrg File: libgomp.info,  Node: omp_get_num_procs,  Next: omp_get_num_teams,  Prev: omp_get_num_devices,  Up: Runtime Library Routines
    440   1.5  mrg 
    441  1.11  mrg 2.12 'omp_get_num_procs' - Number of processors online
    442   1.5  mrg ======================================================
    443   1.5  mrg 
    444   1.5  mrg _Description_:
    445   1.5  mrg      Returns the number of processors online on that device.
    446   1.1  mrg 
    447   1.1  mrg _C/C++_:
    448  1.11  mrg      _Prototype_:   'int omp_get_num_procs(void);'
    449   1.1  mrg 
    450   1.1  mrg _Fortran_:
    451  1.11  mrg      _Interface_:   'integer function omp_get_num_procs()'
    452   1.1  mrg 
    453   1.1  mrg _Reference_:
    454   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.5.
    455   1.1  mrg 
    456   1.1  mrg 
    457   1.5  mrg File: libgomp.info,  Node: omp_get_num_teams,  Next: omp_get_num_threads,  Prev: omp_get_num_procs,  Up: Runtime Library Routines
    458   1.1  mrg 
    459  1.11  mrg 2.13 'omp_get_num_teams' - Number of teams
    460   1.5  mrg ==========================================
    461   1.1  mrg 
    462   1.1  mrg _Description_:
    463   1.5  mrg      Returns the number of teams in the current team region.
    464   1.5  mrg 
    465   1.5  mrg _C/C++_:
    466  1.11  mrg      _Prototype_:   'int omp_get_num_teams(void);'
    467   1.5  mrg 
    468   1.5  mrg _Fortran_:
    469  1.11  mrg      _Interface_:   'integer function omp_get_num_teams()'
    470   1.5  mrg 
    471   1.5  mrg _Reference_:
    472   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.32.
    473   1.5  mrg 
    474   1.5  mrg 
    475   1.5  mrg File: libgomp.info,  Node: omp_get_num_threads,  Next: omp_get_proc_bind,  Prev: omp_get_num_teams,  Up: Runtime Library Routines
    476   1.5  mrg 
    477  1.11  mrg 2.14 'omp_get_num_threads' - Size of the active team
    478   1.5  mrg ====================================================
    479   1.5  mrg 
    480   1.5  mrg _Description_:
    481  1.11  mrg      Returns the number of threads in the current team.  In a sequential
    482  1.11  mrg      section of the program 'omp_get_num_threads' returns 1.
    483   1.1  mrg 
    484   1.1  mrg      The default team size may be initialized at startup by the
    485  1.11  mrg      'OMP_NUM_THREADS' environment variable.  At runtime, the size of
    486  1.11  mrg      the current team may be set either by the 'NUM_THREADS' clause or
    487  1.11  mrg      by 'omp_set_num_threads'.  If none of the above were used to define
    488  1.11  mrg      a specific value and 'OMP_DYNAMIC' is disabled, one thread per CPU
    489  1.11  mrg      online is used.
    490   1.1  mrg 
    491   1.1  mrg _C/C++_:
    492  1.11  mrg      _Prototype_:   'int omp_get_num_threads(void);'
    493   1.1  mrg 
    494   1.1  mrg _Fortran_:
    495  1.11  mrg      _Interface_:   'integer function omp_get_num_threads()'
    496   1.1  mrg 
    497   1.1  mrg _See also_:
    498   1.1  mrg      *note omp_get_max_threads::, *note omp_set_num_threads::, *note
    499   1.1  mrg      OMP_NUM_THREADS::
    500   1.1  mrg 
    501   1.1  mrg _Reference_:
    502   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.2.
    503   1.5  mrg 
    504   1.5  mrg 
    505   1.5  mrg File: libgomp.info,  Node: omp_get_proc_bind,  Next: omp_get_schedule,  Prev: omp_get_num_threads,  Up: Runtime Library Routines
    506   1.5  mrg 
    507  1.11  mrg 2.15 'omp_get_proc_bind' - Whether theads may be moved between CPUs
    508   1.5  mrg ===================================================================
    509   1.5  mrg 
    510   1.5  mrg _Description_:
    511  1.11  mrg      This functions returns the currently active thread affinity policy,
    512  1.11  mrg      which is set via 'OMP_PROC_BIND'.  Possible values are
    513  1.11  mrg      'omp_proc_bind_false', 'omp_proc_bind_true',
    514  1.11  mrg      'omp_proc_bind_master', 'omp_proc_bind_close' and
    515  1.11  mrg      'omp_proc_bind_spread'.
    516   1.5  mrg 
    517   1.5  mrg _C/C++_:
    518  1.11  mrg      _Prototype_:   'omp_proc_bind_t omp_get_proc_bind(void);'
    519   1.5  mrg 
    520   1.5  mrg _Fortran_:
    521  1.11  mrg      _Interface_:   'integer(kind=omp_proc_bind_kind) function
    522  1.11  mrg                     omp_get_proc_bind()'
    523   1.5  mrg 
    524   1.5  mrg _See also_:
    525   1.5  mrg      *note OMP_PROC_BIND::, *note OMP_PLACES::, *note
    526   1.5  mrg      GOMP_CPU_AFFINITY::,
    527   1.5  mrg 
    528   1.5  mrg _Reference_:
    529   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.22.
    530   1.1  mrg 
    531   1.1  mrg 
    532   1.5  mrg File: libgomp.info,  Node: omp_get_schedule,  Next: omp_get_team_num,  Prev: omp_get_proc_bind,  Up: Runtime Library Routines
    533   1.1  mrg 
    534  1.11  mrg 2.16 'omp_get_schedule' - Obtain the runtime scheduling method
    535   1.1  mrg ==============================================================
    536   1.1  mrg 
    537   1.1  mrg _Description_:
    538   1.5  mrg      Obtain the runtime scheduling method.  The KIND argument will be
    539  1.11  mrg      set to the value 'omp_sched_static', 'omp_sched_dynamic',
    540  1.11  mrg      'omp_sched_guided' or 'omp_sched_auto'.  The second argument,
    541   1.7  mrg      CHUNK_SIZE, is set to the chunk size.
    542   1.1  mrg 
    543   1.1  mrg _C/C++_
    544  1.11  mrg      _Prototype_:   'void omp_get_schedule(omp_sched_t *kind, int
    545  1.11  mrg                     *chunk_size);'
    546   1.1  mrg 
    547   1.1  mrg _Fortran_:
    548  1.11  mrg      _Interface_:   'subroutine omp_get_schedule(kind, chunk_size)'
    549  1.11  mrg                     'integer(kind=omp_sched_kind) kind'
    550  1.11  mrg                     'integer chunk_size'
    551   1.1  mrg 
    552   1.1  mrg _See also_:
    553   1.1  mrg      *note omp_set_schedule::, *note OMP_SCHEDULE::
    554   1.1  mrg 
    555   1.1  mrg _Reference_:
    556   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.13.
    557   1.5  mrg 
    558   1.5  mrg 
    559   1.5  mrg File: libgomp.info,  Node: omp_get_team_num,  Next: omp_get_team_size,  Prev: omp_get_schedule,  Up: Runtime Library Routines
    560   1.5  mrg 
    561  1.11  mrg 2.17 'omp_get_team_num' - Get team number
    562   1.5  mrg =========================================
    563   1.5  mrg 
    564   1.5  mrg _Description_:
    565   1.5  mrg      Returns the team number of the calling thread.
    566   1.5  mrg 
    567   1.5  mrg _C/C++_:
    568  1.11  mrg      _Prototype_:   'int omp_get_team_num(void);'
    569   1.5  mrg 
    570   1.5  mrg _Fortran_:
    571  1.11  mrg      _Interface_:   'integer function omp_get_team_num()'
    572   1.5  mrg 
    573   1.5  mrg _Reference_:
    574   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.33.
    575   1.1  mrg 
    576   1.1  mrg 
    577   1.5  mrg File: libgomp.info,  Node: omp_get_team_size,  Next: omp_get_thread_limit,  Prev: omp_get_team_num,  Up: Runtime Library Routines
    578   1.1  mrg 
    579  1.11  mrg 2.18 'omp_get_team_size' - Number of threads in a team
    580   1.1  mrg ======================================================
    581   1.1  mrg 
    582   1.1  mrg _Description_:
    583   1.1  mrg      This function returns the number of threads in a thread team to
    584   1.5  mrg      which either the current thread or its ancestor belongs.  For
    585  1.11  mrg      values of LEVEL outside zero to 'omp_get_level', -1 is returned; if
    586  1.11  mrg      LEVEL is zero, 1 is returned, and for 'omp_get_level', the result
    587  1.11  mrg      is identical to 'omp_get_num_threads'.
    588   1.1  mrg 
    589   1.1  mrg _C/C++_:
    590  1.11  mrg      _Prototype_:   'int omp_get_team_size(int level);'
    591   1.1  mrg 
    592   1.1  mrg _Fortran_:
    593  1.11  mrg      _Interface_:   'integer function omp_get_team_size(level)'
    594  1.11  mrg                     'integer level'
    595   1.1  mrg 
    596   1.1  mrg _See also_:
    597   1.1  mrg      *note omp_get_num_threads::, *note omp_get_level::, *note
    598   1.1  mrg      omp_get_ancestor_thread_num::
    599   1.1  mrg 
    600   1.1  mrg _Reference_:
    601   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.19.
    602   1.1  mrg 
    603   1.1  mrg 
    604   1.1  mrg File: libgomp.info,  Node: omp_get_thread_limit,  Next: omp_get_thread_num,  Prev: omp_get_team_size,  Up: Runtime Library Routines
    605   1.1  mrg 
    606  1.11  mrg 2.19 'omp_get_thread_limit' - Maximum number of threads
    607   1.1  mrg =======================================================
    608   1.1  mrg 
    609   1.1  mrg _Description_:
    610   1.3  mrg      Return the maximum number of threads of the program.
    611   1.1  mrg 
    612   1.1  mrg _C/C++_:
    613  1.11  mrg      _Prototype_:   'int omp_get_thread_limit(void);'
    614   1.1  mrg 
    615   1.1  mrg _Fortran_:
    616  1.11  mrg      _Interface_:   'integer function omp_get_thread_limit()'
    617   1.1  mrg 
    618   1.1  mrg _See also_:
    619   1.1  mrg      *note omp_get_max_threads::, *note OMP_THREAD_LIMIT::
    620   1.1  mrg 
    621   1.1  mrg _Reference_:
    622   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.14.
    623   1.1  mrg 
    624   1.1  mrg 
    625   1.1  mrg File: libgomp.info,  Node: omp_get_thread_num,  Next: omp_in_parallel,  Prev: omp_get_thread_limit,  Up: Runtime Library Routines
    626   1.1  mrg 
    627  1.11  mrg 2.20 'omp_get_thread_num' - Current thread ID
    628   1.1  mrg =============================================
    629   1.1  mrg 
    630   1.1  mrg _Description_:
    631   1.3  mrg      Returns a unique thread identification number within the current
    632  1.11  mrg      team.  In a sequential parts of the program, 'omp_get_thread_num'
    633  1.11  mrg      always returns 0.  In parallel regions the return value varies from
    634  1.11  mrg      0 to 'omp_get_num_threads'-1 inclusive.  The return value of the
    635  1.11  mrg      master thread of a team is always 0.
    636   1.1  mrg 
    637   1.1  mrg _C/C++_:
    638  1.11  mrg      _Prototype_:   'int omp_get_thread_num(void);'
    639   1.1  mrg 
    640   1.1  mrg _Fortran_:
    641  1.11  mrg      _Interface_:   'integer function omp_get_thread_num()'
    642   1.1  mrg 
    643   1.1  mrg _See also_:
    644   1.1  mrg      *note omp_get_num_threads::, *note omp_get_ancestor_thread_num::
    645   1.1  mrg 
    646   1.1  mrg _Reference_:
    647   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.4.
    648   1.1  mrg 
    649   1.1  mrg 
    650   1.3  mrg File: libgomp.info,  Node: omp_in_parallel,  Next: omp_in_final,  Prev: omp_get_thread_num,  Up: Runtime Library Routines
    651   1.1  mrg 
    652  1.11  mrg 2.21 'omp_in_parallel' - Whether a parallel region is active
    653   1.1  mrg ============================================================
    654   1.1  mrg 
    655   1.1  mrg _Description_:
    656  1.11  mrg      This function returns 'true' if currently running in parallel,
    657  1.11  mrg      'false' otherwise.  Here, 'true' and 'false' represent their
    658   1.1  mrg      language-specific counterparts.
    659   1.1  mrg 
    660   1.1  mrg _C/C++_:
    661  1.11  mrg      _Prototype_:   'int omp_in_parallel(void);'
    662   1.1  mrg 
    663   1.1  mrg _Fortran_:
    664  1.11  mrg      _Interface_:   'logical function omp_in_parallel()'
    665   1.1  mrg 
    666   1.1  mrg _Reference_:
    667   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.6.
    668   1.1  mrg 
    669   1.1  mrg 
    670   1.5  mrg File: libgomp.info,  Node: omp_in_final,  Next: omp_is_initial_device,  Prev: omp_in_parallel,  Up: Runtime Library Routines
    671   1.3  mrg 
    672  1.11  mrg 2.22 'omp_in_final' - Whether in final or included task region
    673   1.3  mrg ==============================================================
    674   1.3  mrg 
    675   1.3  mrg _Description_:
    676  1.11  mrg      This function returns 'true' if currently running in a final or
    677  1.11  mrg      included task region, 'false' otherwise.  Here, 'true' and 'false'
    678   1.3  mrg      represent their language-specific counterparts.
    679   1.3  mrg 
    680   1.3  mrg _C/C++_:
    681  1.11  mrg      _Prototype_:   'int omp_in_final(void);'
    682   1.1  mrg 
    683   1.3  mrg _Fortran_:
    684  1.11  mrg      _Interface_:   'logical function omp_in_final()'
    685   1.3  mrg 
    686   1.3  mrg _Reference_:
    687   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.21.
    688   1.5  mrg 
    689   1.5  mrg 
    690   1.5  mrg File: libgomp.info,  Node: omp_is_initial_device,  Next: omp_set_default_device,  Prev: omp_in_final,  Up: Runtime Library Routines
    691   1.5  mrg 
    692  1.11  mrg 2.23 'omp_is_initial_device' - Whether executing on the host device
    693   1.5  mrg ===================================================================
    694   1.5  mrg 
    695   1.5  mrg _Description_:
    696  1.11  mrg      This function returns 'true' if currently running on the host
    697  1.11  mrg      device, 'false' otherwise.  Here, 'true' and 'false' represent
    698   1.5  mrg      their language-specific counterparts.
    699   1.5  mrg 
    700   1.5  mrg _C/C++_:
    701  1.11  mrg      _Prototype_:   'int omp_is_initial_device(void);'
    702   1.5  mrg 
    703   1.5  mrg _Fortran_:
    704  1.11  mrg      _Interface_:   'logical function omp_is_initial_device()'
    705   1.5  mrg 
    706   1.5  mrg _Reference_:
    707   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.34.
    708   1.5  mrg 
    709   1.5  mrg 
    710   1.5  mrg File: libgomp.info,  Node: omp_set_default_device,  Next: omp_set_dynamic,  Prev: omp_is_initial_device,  Up: Runtime Library Routines
    711   1.5  mrg 
    712  1.11  mrg 2.24 'omp_set_default_device' - Set the default device for target regions
    713   1.5  mrg =========================================================================
    714   1.5  mrg 
    715   1.5  mrg _Description_:
    716   1.5  mrg      Set the default device for target regions without device clause.
    717   1.5  mrg      The argument shall be a nonnegative device number.
    718   1.5  mrg 
    719   1.5  mrg _C/C++_:
    720  1.11  mrg      _Prototype_:   'void omp_set_default_device(int device_num);'
    721   1.5  mrg 
    722   1.5  mrg _Fortran_:
    723  1.11  mrg      _Interface_:   'subroutine omp_set_default_device(device_num)'
    724  1.11  mrg                     'integer device_num'
    725   1.5  mrg 
    726   1.5  mrg _See also_:
    727   1.5  mrg      *note OMP_DEFAULT_DEVICE::, *note omp_get_default_device::
    728   1.5  mrg 
    729   1.5  mrg _Reference_:
    730   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.29.
    731   1.3  mrg 
    732   1.3  mrg 
    733   1.5  mrg File: libgomp.info,  Node: omp_set_dynamic,  Next: omp_set_max_active_levels,  Prev: omp_set_default_device,  Up: Runtime Library Routines
    734   1.3  mrg 
    735  1.11  mrg 2.25 'omp_set_dynamic' - Enable/disable dynamic teams
    736   1.1  mrg =====================================================
    737   1.1  mrg 
    738   1.1  mrg _Description_:
    739   1.1  mrg      Enable or disable the dynamic adjustment of the number of threads
    740   1.5  mrg      within a team.  The function takes the language-specific equivalent
    741  1.11  mrg      of 'true' and 'false', where 'true' enables dynamic adjustment of
    742  1.11  mrg      team sizes and 'false' disables it.
    743   1.1  mrg 
    744   1.1  mrg _C/C++_:
    745  1.11  mrg      _Prototype_:   'void omp_set_dynamic(int dynamic_threads);'
    746   1.1  mrg 
    747   1.1  mrg _Fortran_:
    748  1.11  mrg      _Interface_:   'subroutine omp_set_dynamic(dynamic_threads)'
    749  1.11  mrg                     'logical, intent(in) :: dynamic_threads'
    750   1.1  mrg 
    751   1.1  mrg _See also_:
    752   1.1  mrg      *note OMP_DYNAMIC::, *note omp_get_dynamic::
    753   1.1  mrg 
    754   1.1  mrg _Reference_:
    755   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.7.
    756   1.1  mrg 
    757   1.1  mrg 
    758   1.1  mrg File: libgomp.info,  Node: omp_set_max_active_levels,  Next: omp_set_nested,  Prev: omp_set_dynamic,  Up: Runtime Library Routines
    759   1.1  mrg 
    760  1.11  mrg 2.26 'omp_set_max_active_levels' - Limits the number of active parallel regions
    761   1.1  mrg ===============================================================================
    762   1.1  mrg 
    763   1.1  mrg _Description_:
    764   1.3  mrg      This function limits the maximum allowed number of nested, active
    765   1.3  mrg      parallel regions.
    766   1.1  mrg 
    767   1.1  mrg _C/C++_
    768  1.11  mrg      _Prototype_:   'void omp_set_max_active_levels(int max_levels);'
    769   1.1  mrg 
    770   1.1  mrg _Fortran_:
    771  1.11  mrg      _Interface_:   'subroutine omp_set_max_active_levels(max_levels)'
    772  1.11  mrg                     'integer max_levels'
    773   1.1  mrg 
    774   1.1  mrg _See also_:
    775   1.1  mrg      *note omp_get_max_active_levels::, *note omp_get_active_level::
    776   1.1  mrg 
    777   1.1  mrg _Reference_:
    778   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.15.
    779   1.1  mrg 
    780   1.1  mrg 
    781   1.1  mrg File: libgomp.info,  Node: omp_set_nested,  Next: omp_set_num_threads,  Prev: omp_set_max_active_levels,  Up: Runtime Library Routines
    782   1.1  mrg 
    783  1.11  mrg 2.27 'omp_set_nested' - Enable/disable nested parallel regions
    784   1.1  mrg ==============================================================
    785   1.1  mrg 
    786   1.1  mrg _Description_:
    787   1.1  mrg      Enable or disable nested parallel regions, i.e., whether team
    788   1.5  mrg      members are allowed to create new teams.  The function takes the
    789  1.11  mrg      language-specific equivalent of 'true' and 'false', where 'true'
    790  1.11  mrg      enables dynamic adjustment of team sizes and 'false' disables it.
    791   1.1  mrg 
    792   1.1  mrg _C/C++_:
    793  1.11  mrg      _Prototype_:   'void omp_set_nested(int nested);'
    794   1.1  mrg 
    795   1.1  mrg _Fortran_:
    796  1.11  mrg      _Interface_:   'subroutine omp_set_nested(nested)'
    797  1.11  mrg                     'logical, intent(in) :: nested'
    798   1.1  mrg 
    799   1.1  mrg _See also_:
    800   1.1  mrg      *note OMP_NESTED::, *note omp_get_nested::
    801   1.1  mrg 
    802   1.1  mrg _Reference_:
    803   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.10.
    804   1.1  mrg 
    805   1.1  mrg 
    806   1.1  mrg File: libgomp.info,  Node: omp_set_num_threads,  Next: omp_set_schedule,  Prev: omp_set_nested,  Up: Runtime Library Routines
    807   1.1  mrg 
    808  1.11  mrg 2.28 'omp_set_num_threads' - Set upper team size limit
    809   1.1  mrg ======================================================
    810   1.1  mrg 
    811   1.1  mrg _Description_:
    812   1.1  mrg      Specifies the number of threads used by default in subsequent
    813  1.11  mrg      parallel sections, if those do not specify a 'num_threads' clause.
    814  1.11  mrg      The argument of 'omp_set_num_threads' shall be a positive integer.
    815   1.1  mrg 
    816   1.1  mrg _C/C++_:
    817  1.11  mrg      _Prototype_:   'void omp_set_num_threads(int num_threads);'
    818   1.1  mrg 
    819   1.1  mrg _Fortran_:
    820  1.11  mrg      _Interface_:   'subroutine omp_set_num_threads(num_threads)'
    821  1.11  mrg                     'integer, intent(in) :: num_threads'
    822   1.1  mrg 
    823   1.1  mrg _See also_:
    824   1.1  mrg      *note OMP_NUM_THREADS::, *note omp_get_num_threads::, *note
    825   1.1  mrg      omp_get_max_threads::
    826   1.1  mrg 
    827   1.1  mrg _Reference_:
    828   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.1.
    829   1.1  mrg 
    830   1.1  mrg 
    831   1.1  mrg File: libgomp.info,  Node: omp_set_schedule,  Next: omp_init_lock,  Prev: omp_set_num_threads,  Up: Runtime Library Routines
    832   1.1  mrg 
    833  1.11  mrg 2.29 'omp_set_schedule' - Set the runtime scheduling method
    834   1.1  mrg ===========================================================
    835   1.1  mrg 
    836   1.1  mrg _Description_:
    837   1.5  mrg      Sets the runtime scheduling method.  The KIND argument can have the
    838  1.11  mrg      value 'omp_sched_static', 'omp_sched_dynamic', 'omp_sched_guided'
    839  1.11  mrg      or 'omp_sched_auto'.  Except for 'omp_sched_auto', the chunk size
    840   1.7  mrg      is set to the value of CHUNK_SIZE if positive, or to the default
    841  1.11  mrg      value if zero or negative.  For 'omp_sched_auto' the CHUNK_SIZE
    842   1.1  mrg      argument is ignored.
    843   1.1  mrg 
    844   1.1  mrg _C/C++_
    845  1.11  mrg      _Prototype_:   'void omp_set_schedule(omp_sched_t kind, int
    846  1.11  mrg                     chunk_size);'
    847   1.1  mrg 
    848   1.1  mrg _Fortran_:
    849  1.11  mrg      _Interface_:   'subroutine omp_set_schedule(kind, chunk_size)'
    850  1.11  mrg                     'integer(kind=omp_sched_kind) kind'
    851  1.11  mrg                     'integer chunk_size'
    852   1.1  mrg 
    853   1.1  mrg _See also_:
    854   1.1  mrg      *note omp_get_schedule:: *note OMP_SCHEDULE::
    855   1.1  mrg 
    856   1.1  mrg _Reference_:
    857   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.2.12.
    858   1.1  mrg 
    859   1.1  mrg 
    860   1.1  mrg File: libgomp.info,  Node: omp_init_lock,  Next: omp_set_lock,  Prev: omp_set_schedule,  Up: Runtime Library Routines
    861   1.1  mrg 
    862  1.11  mrg 2.30 'omp_init_lock' - Initialize simple lock
    863   1.1  mrg =============================================
    864   1.1  mrg 
    865   1.1  mrg _Description_:
    866   1.3  mrg      Initialize a simple lock.  After initialization, the lock is in an
    867   1.1  mrg      unlocked state.
    868   1.1  mrg 
    869   1.1  mrg _C/C++_:
    870  1.11  mrg      _Prototype_:   'void omp_init_lock(omp_lock_t *lock);'
    871   1.1  mrg 
    872   1.1  mrg _Fortran_:
    873  1.11  mrg      _Interface_:   'subroutine omp_init_lock(svar)'
    874  1.11  mrg                     'integer(omp_lock_kind), intent(out) :: svar'
    875   1.1  mrg 
    876   1.1  mrg _See also_:
    877   1.1  mrg      *note omp_destroy_lock::
    878   1.1  mrg 
    879   1.1  mrg _Reference_:
    880   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.1.
    881   1.1  mrg 
    882   1.1  mrg 
    883   1.1  mrg File: libgomp.info,  Node: omp_set_lock,  Next: omp_test_lock,  Prev: omp_init_lock,  Up: Runtime Library Routines
    884   1.1  mrg 
    885  1.11  mrg 2.31 'omp_set_lock' - Wait for and set simple lock
    886   1.1  mrg ==================================================
    887   1.1  mrg 
    888   1.1  mrg _Description_:
    889  1.11  mrg      Before setting a simple lock, the lock variable must be initialized
    890  1.11  mrg      by 'omp_init_lock'.  The calling thread is blocked until the lock
    891  1.11  mrg      is available.  If the lock is already held by the current thread, a
    892  1.11  mrg      deadlock occurs.
    893   1.1  mrg 
    894   1.1  mrg _C/C++_:
    895  1.11  mrg      _Prototype_:   'void omp_set_lock(omp_lock_t *lock);'
    896   1.1  mrg 
    897   1.1  mrg _Fortran_:
    898  1.11  mrg      _Interface_:   'subroutine omp_set_lock(svar)'
    899  1.11  mrg                     'integer(omp_lock_kind), intent(inout) :: svar'
    900   1.1  mrg 
    901   1.1  mrg _See also_:
    902   1.1  mrg      *note omp_init_lock::, *note omp_test_lock::, *note
    903   1.1  mrg      omp_unset_lock::
    904   1.1  mrg 
    905   1.1  mrg _Reference_:
    906   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.4.
    907   1.1  mrg 
    908   1.1  mrg 
    909   1.1  mrg File: libgomp.info,  Node: omp_test_lock,  Next: omp_unset_lock,  Prev: omp_set_lock,  Up: Runtime Library Routines
    910   1.1  mrg 
    911  1.11  mrg 2.32 'omp_test_lock' - Test and set simple lock if available
    912   1.1  mrg ============================================================
    913   1.1  mrg 
    914   1.1  mrg _Description_:
    915  1.11  mrg      Before setting a simple lock, the lock variable must be initialized
    916  1.11  mrg      by 'omp_init_lock'.  Contrary to 'omp_set_lock', 'omp_test_lock'
    917  1.11  mrg      does not block if the lock is not available.  This function returns
    918  1.11  mrg      'true' upon success, 'false' otherwise.  Here, 'true' and 'false'
    919  1.11  mrg      represent their language-specific counterparts.
    920   1.1  mrg 
    921   1.1  mrg _C/C++_:
    922  1.11  mrg      _Prototype_:   'int omp_test_lock(omp_lock_t *lock);'
    923   1.1  mrg 
    924   1.1  mrg _Fortran_:
    925  1.11  mrg      _Interface_:   'logical function omp_test_lock(svar)'
    926  1.11  mrg                     'integer(omp_lock_kind), intent(inout) :: svar'
    927   1.1  mrg 
    928   1.1  mrg _See also_:
    929   1.1  mrg      *note omp_init_lock::, *note omp_set_lock::, *note omp_set_lock::
    930   1.1  mrg 
    931   1.1  mrg _Reference_:
    932   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.6.
    933   1.1  mrg 
    934   1.1  mrg 
    935   1.1  mrg File: libgomp.info,  Node: omp_unset_lock,  Next: omp_destroy_lock,  Prev: omp_test_lock,  Up: Runtime Library Routines
    936   1.1  mrg 
    937  1.11  mrg 2.33 'omp_unset_lock' - Unset simple lock
    938   1.1  mrg =========================================
    939   1.1  mrg 
    940   1.1  mrg _Description_:
    941   1.1  mrg      A simple lock about to be unset must have been locked by
    942  1.11  mrg      'omp_set_lock' or 'omp_test_lock' before.  In addition, the lock
    943  1.11  mrg      must be held by the thread calling 'omp_unset_lock'.  Then, the
    944  1.11  mrg      lock becomes unlocked.  If one or more threads attempted to set the
    945  1.11  mrg      lock before, one of them is chosen to, again, set the lock to
    946   1.1  mrg      itself.
    947   1.1  mrg 
    948   1.1  mrg _C/C++_:
    949  1.11  mrg      _Prototype_:   'void omp_unset_lock(omp_lock_t *lock);'
    950   1.1  mrg 
    951   1.1  mrg _Fortran_:
    952  1.11  mrg      _Interface_:   'subroutine omp_unset_lock(svar)'
    953  1.11  mrg                     'integer(omp_lock_kind), intent(inout) :: svar'
    954   1.1  mrg 
    955   1.1  mrg _See also_:
    956   1.1  mrg      *note omp_set_lock::, *note omp_test_lock::
    957   1.1  mrg 
    958   1.1  mrg _Reference_:
    959   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.5.
    960   1.1  mrg 
    961   1.1  mrg 
    962   1.1  mrg File: libgomp.info,  Node: omp_destroy_lock,  Next: omp_init_nest_lock,  Prev: omp_unset_lock,  Up: Runtime Library Routines
    963   1.1  mrg 
    964  1.11  mrg 2.34 'omp_destroy_lock' - Destroy simple lock
    965   1.1  mrg =============================================
    966   1.1  mrg 
    967   1.1  mrg _Description_:
    968   1.5  mrg      Destroy a simple lock.  In order to be destroyed, a simple lock
    969   1.1  mrg      must be in the unlocked state.
    970   1.1  mrg 
    971   1.1  mrg _C/C++_:
    972  1.11  mrg      _Prototype_:   'void omp_destroy_lock(omp_lock_t *lock);'
    973   1.1  mrg 
    974   1.1  mrg _Fortran_:
    975  1.11  mrg      _Interface_:   'subroutine omp_destroy_lock(svar)'
    976  1.11  mrg                     'integer(omp_lock_kind), intent(inout) :: svar'
    977   1.1  mrg 
    978   1.1  mrg _See also_:
    979   1.1  mrg      *note omp_init_lock::
    980   1.1  mrg 
    981   1.1  mrg _Reference_:
    982   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.3.
    983   1.1  mrg 
    984   1.1  mrg 
    985   1.1  mrg File: libgomp.info,  Node: omp_init_nest_lock,  Next: omp_set_nest_lock,  Prev: omp_destroy_lock,  Up: Runtime Library Routines
    986   1.1  mrg 
    987  1.11  mrg 2.35 'omp_init_nest_lock' - Initialize nested lock
    988   1.1  mrg ==================================================
    989   1.1  mrg 
    990   1.1  mrg _Description_:
    991   1.3  mrg      Initialize a nested lock.  After initialization, the lock is in an
    992   1.1  mrg      unlocked state and the nesting count is set to zero.
    993   1.1  mrg 
    994   1.1  mrg _C/C++_:
    995  1.11  mrg      _Prototype_:   'void omp_init_nest_lock(omp_nest_lock_t *lock);'
    996   1.1  mrg 
    997   1.1  mrg _Fortran_:
    998  1.11  mrg      _Interface_:   'subroutine omp_init_nest_lock(nvar)'
    999  1.11  mrg                     'integer(omp_nest_lock_kind), intent(out) :: nvar'
   1000   1.1  mrg 
   1001   1.1  mrg _See also_:
   1002   1.1  mrg      *note omp_destroy_nest_lock::
   1003   1.1  mrg 
   1004   1.1  mrg _Reference_:
   1005   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.1.
   1006   1.1  mrg 
   1007   1.1  mrg 
   1008   1.1  mrg File: libgomp.info,  Node: omp_set_nest_lock,  Next: omp_test_nest_lock,  Prev: omp_init_nest_lock,  Up: Runtime Library Routines
   1009   1.1  mrg 
   1010  1.11  mrg 2.36 'omp_set_nest_lock' - Wait for and set nested lock
   1011   1.1  mrg =======================================================
   1012   1.1  mrg 
   1013   1.1  mrg _Description_:
   1014  1.11  mrg      Before setting a nested lock, the lock variable must be initialized
   1015  1.11  mrg      by 'omp_init_nest_lock'.  The calling thread is blocked until the
   1016  1.11  mrg      lock is available.  If the lock is already held by the current
   1017  1.11  mrg      thread, the nesting count for the lock is incremented.
   1018   1.1  mrg 
   1019   1.1  mrg _C/C++_:
   1020  1.11  mrg      _Prototype_:   'void omp_set_nest_lock(omp_nest_lock_t *lock);'
   1021   1.1  mrg 
   1022   1.1  mrg _Fortran_:
   1023  1.11  mrg      _Interface_:   'subroutine omp_set_nest_lock(nvar)'
   1024  1.11  mrg                     'integer(omp_nest_lock_kind), intent(inout) :: nvar'
   1025   1.1  mrg 
   1026   1.1  mrg _See also_:
   1027   1.1  mrg      *note omp_init_nest_lock::, *note omp_unset_nest_lock::
   1028   1.1  mrg 
   1029   1.1  mrg _Reference_:
   1030   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.4.
   1031   1.1  mrg 
   1032   1.1  mrg 
   1033   1.1  mrg File: libgomp.info,  Node: omp_test_nest_lock,  Next: omp_unset_nest_lock,  Prev: omp_set_nest_lock,  Up: Runtime Library Routines
   1034   1.1  mrg 
   1035  1.11  mrg 2.37 'omp_test_nest_lock' - Test and set nested lock if available
   1036   1.1  mrg =================================================================
   1037   1.1  mrg 
   1038   1.1  mrg _Description_:
   1039  1.11  mrg      Before setting a nested lock, the lock variable must be initialized
   1040  1.11  mrg      by 'omp_init_nest_lock'.  Contrary to 'omp_set_nest_lock',
   1041  1.11  mrg      'omp_test_nest_lock' does not block if the lock is not available.
   1042  1.11  mrg      If the lock is already held by the current thread, the new nesting
   1043  1.11  mrg      count is returned.  Otherwise, the return value equals zero.
   1044   1.1  mrg 
   1045   1.1  mrg _C/C++_:
   1046  1.11  mrg      _Prototype_:   'int omp_test_nest_lock(omp_nest_lock_t *lock);'
   1047   1.1  mrg 
   1048   1.1  mrg _Fortran_:
   1049  1.11  mrg      _Interface_:   'logical function omp_test_nest_lock(nvar)'
   1050  1.11  mrg                     'integer(omp_nest_lock_kind), intent(inout) :: nvar'
   1051   1.1  mrg 
   1052   1.1  mrg _See also_:
   1053   1.1  mrg      *note omp_init_lock::, *note omp_set_lock::, *note omp_set_lock::
   1054   1.1  mrg 
   1055   1.1  mrg _Reference_:
   1056   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.6.
   1057   1.1  mrg 
   1058   1.1  mrg 
   1059   1.1  mrg File: libgomp.info,  Node: omp_unset_nest_lock,  Next: omp_destroy_nest_lock,  Prev: omp_test_nest_lock,  Up: Runtime Library Routines
   1060   1.1  mrg 
   1061  1.11  mrg 2.38 'omp_unset_nest_lock' - Unset nested lock
   1062   1.1  mrg ==============================================
   1063   1.1  mrg 
   1064   1.1  mrg _Description_:
   1065   1.1  mrg      A nested lock about to be unset must have been locked by
   1066  1.11  mrg      'omp_set_nested_lock' or 'omp_test_nested_lock' before.  In
   1067   1.1  mrg      addition, the lock must be held by the thread calling
   1068  1.11  mrg      'omp_unset_nested_lock'.  If the nesting count drops to zero, the
   1069   1.5  mrg      lock becomes unlocked.  If one ore more threads attempted to set
   1070   1.3  mrg      the lock before, one of them is chosen to, again, set the lock to
   1071   1.1  mrg      itself.
   1072   1.1  mrg 
   1073   1.1  mrg _C/C++_:
   1074  1.11  mrg      _Prototype_:   'void omp_unset_nest_lock(omp_nest_lock_t *lock);'
   1075   1.1  mrg 
   1076   1.1  mrg _Fortran_:
   1077  1.11  mrg      _Interface_:   'subroutine omp_unset_nest_lock(nvar)'
   1078  1.11  mrg                     'integer(omp_nest_lock_kind), intent(inout) :: nvar'
   1079   1.1  mrg 
   1080   1.1  mrg _See also_:
   1081   1.1  mrg      *note omp_set_nest_lock::
   1082   1.1  mrg 
   1083   1.1  mrg _Reference_:
   1084   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.5.
   1085   1.1  mrg 
   1086   1.1  mrg 
   1087   1.1  mrg File: libgomp.info,  Node: omp_destroy_nest_lock,  Next: omp_get_wtick,  Prev: omp_unset_nest_lock,  Up: Runtime Library Routines
   1088   1.1  mrg 
   1089  1.11  mrg 2.39 'omp_destroy_nest_lock' - Destroy nested lock
   1090   1.1  mrg ==================================================
   1091   1.1  mrg 
   1092   1.1  mrg _Description_:
   1093   1.5  mrg      Destroy a nested lock.  In order to be destroyed, a nested lock
   1094   1.1  mrg      must be in the unlocked state and its nesting count must equal
   1095   1.1  mrg      zero.
   1096   1.1  mrg 
   1097   1.1  mrg _C/C++_:
   1098  1.11  mrg      _Prototype_:   'void omp_destroy_nest_lock(omp_nest_lock_t *);'
   1099   1.1  mrg 
   1100   1.1  mrg _Fortran_:
   1101  1.11  mrg      _Interface_:   'subroutine omp_destroy_nest_lock(nvar)'
   1102  1.11  mrg                     'integer(omp_nest_lock_kind), intent(inout) :: nvar'
   1103   1.1  mrg 
   1104   1.1  mrg _See also_:
   1105   1.1  mrg      *note omp_init_lock::
   1106   1.1  mrg 
   1107   1.1  mrg _Reference_:
   1108   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.3.3.
   1109   1.1  mrg 
   1110   1.1  mrg 
   1111   1.1  mrg File: libgomp.info,  Node: omp_get_wtick,  Next: omp_get_wtime,  Prev: omp_destroy_nest_lock,  Up: Runtime Library Routines
   1112   1.1  mrg 
   1113  1.11  mrg 2.40 'omp_get_wtick' - Get timer precision
   1114   1.1  mrg ==========================================
   1115   1.1  mrg 
   1116   1.1  mrg _Description_:
   1117   1.1  mrg      Gets the timer precision, i.e., the number of seconds between two
   1118   1.1  mrg      successive clock ticks.
   1119   1.1  mrg 
   1120   1.1  mrg _C/C++_:
   1121  1.11  mrg      _Prototype_:   'double omp_get_wtick(void);'
   1122   1.1  mrg 
   1123   1.1  mrg _Fortran_:
   1124  1.11  mrg      _Interface_:   'double precision function omp_get_wtick()'
   1125   1.1  mrg 
   1126   1.1  mrg _See also_:
   1127   1.1  mrg      *note omp_get_wtime::
   1128   1.1  mrg 
   1129   1.1  mrg _Reference_:
   1130   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.4.2.
   1131   1.1  mrg 
   1132   1.1  mrg 
   1133   1.1  mrg File: libgomp.info,  Node: omp_get_wtime,  Prev: omp_get_wtick,  Up: Runtime Library Routines
   1134   1.1  mrg 
   1135  1.11  mrg 2.41 'omp_get_wtime' - Elapsed wall clock time
   1136   1.1  mrg ==============================================
   1137   1.1  mrg 
   1138   1.1  mrg _Description_:
   1139   1.5  mrg      Elapsed wall clock time in seconds.  The time is measured per
   1140   1.3  mrg      thread, no guarantee can be made that two distinct threads measure
   1141   1.3  mrg      the same time.  Time is measured from some "time in the past",
   1142   1.3  mrg      which is an arbitrary time guaranteed not to change during the
   1143   1.3  mrg      execution of the program.
   1144   1.1  mrg 
   1145   1.1  mrg _C/C++_:
   1146  1.11  mrg      _Prototype_:   'double omp_get_wtime(void);'
   1147   1.1  mrg 
   1148   1.1  mrg _Fortran_:
   1149  1.11  mrg      _Interface_:   'double precision function omp_get_wtime()'
   1150   1.1  mrg 
   1151   1.1  mrg _See also_:
   1152   1.1  mrg      *note omp_get_wtick::
   1153   1.1  mrg 
   1154   1.1  mrg _Reference_:
   1155   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 3.4.1.
   1156   1.1  mrg 
   1157   1.1  mrg 
   1158   1.7  mrg File: libgomp.info,  Node: Environment Variables,  Next: Enabling OpenACC,  Prev: Runtime Library Routines,  Up: Top
   1159   1.1  mrg 
   1160   1.1  mrg 3 Environment Variables
   1161   1.1  mrg ***********************
   1162   1.1  mrg 
   1163  1.11  mrg The environment variables which beginning with 'OMP_' are defined by
   1164   1.7  mrg section 4 of the OpenMP specification in version 4.5, while those
   1165  1.11  mrg beginning with 'GOMP_' are GNU extensions.
   1166   1.1  mrg 
   1167   1.1  mrg * Menu:
   1168   1.1  mrg 
   1169   1.7  mrg * OMP_CANCELLATION::        Set whether cancellation is activated
   1170   1.7  mrg * OMP_DISPLAY_ENV::         Show OpenMP version and environment variables
   1171   1.7  mrg * OMP_DEFAULT_DEVICE::      Set the device used in target regions
   1172   1.7  mrg * OMP_DYNAMIC::             Dynamic adjustment of threads
   1173   1.7  mrg * OMP_MAX_ACTIVE_LEVELS::   Set the maximum number of nested parallel regions
   1174   1.7  mrg * OMP_MAX_TASK_PRIORITY::   Set the maximum task priority value
   1175   1.7  mrg * OMP_NESTED::              Nested parallel regions
   1176   1.7  mrg * OMP_NUM_THREADS::         Specifies the number of threads to use
   1177   1.7  mrg * OMP_PROC_BIND::           Whether theads may be moved between CPUs
   1178   1.7  mrg * OMP_PLACES::              Specifies on which CPUs the theads should be placed
   1179   1.7  mrg * OMP_STACKSIZE::           Set default thread stack size
   1180   1.7  mrg * OMP_SCHEDULE::            How threads are scheduled
   1181   1.7  mrg * OMP_THREAD_LIMIT::        Set the maximum number of threads
   1182   1.7  mrg * OMP_WAIT_POLICY::         How waiting threads are handled
   1183   1.7  mrg * GOMP_CPU_AFFINITY::       Bind threads to specific CPUs
   1184   1.7  mrg * GOMP_DEBUG::              Enable debugging output
   1185   1.7  mrg * GOMP_STACKSIZE::          Set default thread stack size
   1186   1.7  mrg * GOMP_SPINCOUNT::          Set the busy-wait spin count
   1187   1.7  mrg * GOMP_RTEMS_THREAD_POOLS:: Set the RTEMS specific thread pools
   1188   1.1  mrg 
   1189   1.1  mrg 
   1190   1.5  mrg File: libgomp.info,  Node: OMP_CANCELLATION,  Next: OMP_DISPLAY_ENV,  Up: Environment Variables
   1191   1.5  mrg 
   1192  1.11  mrg 3.1 'OMP_CANCELLATION' - Set whether cancellation is activated
   1193   1.5  mrg ==============================================================
   1194   1.5  mrg 
   1195   1.5  mrg _Description_:
   1196  1.11  mrg      If set to 'TRUE', the cancellation is activated.  If set to 'FALSE'
   1197  1.11  mrg      or if unset, cancellation is disabled and the 'cancel' construct is
   1198  1.11  mrg      ignored.
   1199   1.5  mrg 
   1200   1.5  mrg _See also_:
   1201   1.5  mrg      *note omp_get_cancellation::
   1202   1.1  mrg 
   1203   1.5  mrg _Reference_:
   1204   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.11
   1205   1.5  mrg 
   1206   1.5  mrg 
   1207   1.5  mrg File: libgomp.info,  Node: OMP_DISPLAY_ENV,  Next: OMP_DEFAULT_DEVICE,  Prev: OMP_CANCELLATION,  Up: Environment Variables
   1208   1.5  mrg 
   1209  1.11  mrg 3.2 'OMP_DISPLAY_ENV' - Show OpenMP version and environment variables
   1210   1.5  mrg =====================================================================
   1211   1.5  mrg 
   1212   1.5  mrg _Description_:
   1213  1.11  mrg      If set to 'TRUE', the OpenMP version number and the values
   1214   1.5  mrg      associated with the OpenMP environment variables are printed to
   1215  1.11  mrg      'stderr'.  If set to 'VERBOSE', it additionally shows the value of
   1216   1.5  mrg      the environment variables which are GNU extensions.  If undefined
   1217  1.11  mrg      or set to 'FALSE', this information will not be shown.
   1218   1.5  mrg 
   1219   1.5  mrg _Reference_:
   1220   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.12
   1221   1.5  mrg 
   1222   1.5  mrg 
   1223   1.5  mrg File: libgomp.info,  Node: OMP_DEFAULT_DEVICE,  Next: OMP_DYNAMIC,  Prev: OMP_DISPLAY_ENV,  Up: Environment Variables
   1224   1.5  mrg 
   1225  1.11  mrg 3.3 'OMP_DEFAULT_DEVICE' - Set the device used in target regions
   1226   1.5  mrg ================================================================
   1227   1.5  mrg 
   1228   1.5  mrg _Description_:
   1229  1.11  mrg      Set to choose the device which is used in a 'target' region, unless
   1230  1.11  mrg      the value is overridden by 'omp_set_default_device' or by a
   1231  1.11  mrg      'device' clause.  The value shall be the nonnegative device number.
   1232  1.11  mrg      If no device with the given device number exists, the code is
   1233  1.11  mrg      executed on the host.  If unset, device number 0 will be used.
   1234   1.5  mrg 
   1235   1.5  mrg _See also_:
   1236   1.5  mrg      *note omp_get_default_device::, *note omp_set_default_device::,
   1237   1.5  mrg 
   1238   1.5  mrg _Reference_:
   1239   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.13
   1240   1.5  mrg 
   1241   1.5  mrg 
   1242   1.5  mrg File: libgomp.info,  Node: OMP_DYNAMIC,  Next: OMP_MAX_ACTIVE_LEVELS,  Prev: OMP_DEFAULT_DEVICE,  Up: Environment Variables
   1243   1.5  mrg 
   1244  1.11  mrg 3.4 'OMP_DYNAMIC' - Dynamic adjustment of threads
   1245   1.1  mrg =================================================
   1246   1.1  mrg 
   1247   1.1  mrg _Description_:
   1248   1.1  mrg      Enable or disable the dynamic adjustment of the number of threads
   1249   1.5  mrg      within a team.  The value of this environment variable shall be
   1250  1.11  mrg      'TRUE' or 'FALSE'.  If undefined, dynamic adjustment is disabled by
   1251  1.11  mrg      default.
   1252   1.1  mrg 
   1253   1.1  mrg _See also_:
   1254   1.1  mrg      *note omp_set_dynamic::
   1255   1.1  mrg 
   1256   1.1  mrg _Reference_:
   1257   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.3
   1258   1.1  mrg 
   1259   1.1  mrg 
   1260   1.7  mrg File: libgomp.info,  Node: OMP_MAX_ACTIVE_LEVELS,  Next: OMP_MAX_TASK_PRIORITY,  Prev: OMP_DYNAMIC,  Up: Environment Variables
   1261   1.1  mrg 
   1262  1.11  mrg 3.5 'OMP_MAX_ACTIVE_LEVELS' - Set the maximum number of nested parallel regions
   1263   1.1  mrg ===============================================================================
   1264   1.1  mrg 
   1265   1.1  mrg _Description_:
   1266   1.3  mrg      Specifies the initial value for the maximum number of nested
   1267   1.5  mrg      parallel regions.  The value of this variable shall be a positive
   1268   1.1  mrg      integer.  If undefined, the number of active levels is unlimited.
   1269   1.1  mrg 
   1270   1.1  mrg _See also_:
   1271   1.1  mrg      *note omp_set_max_active_levels::
   1272   1.1  mrg 
   1273   1.1  mrg _Reference_:
   1274   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.9
   1275   1.1  mrg 
   1276   1.1  mrg 
   1277   1.7  mrg File: libgomp.info,  Node: OMP_MAX_TASK_PRIORITY,  Next: OMP_NESTED,  Prev: OMP_MAX_ACTIVE_LEVELS,  Up: Environment Variables
   1278   1.7  mrg 
   1279  1.11  mrg 3.6 'OMP_MAX_TASK_PRIORITY' - Set the maximum priority
   1280   1.7  mrg ======================================================
   1281   1.7  mrg 
   1282  1.11  mrg number that can be set for a task.
   1283   1.7  mrg _Description_:
   1284  1.11  mrg      Specifies the initial value for the maximum priority value that can
   1285  1.11  mrg      be set for a task.  The value of this variable shall be a
   1286   1.7  mrg      non-negative integer, and zero is allowed.  If undefined, the
   1287   1.7  mrg      default priority is 0.
   1288   1.7  mrg 
   1289   1.7  mrg _See also_:
   1290   1.7  mrg      *note omp_get_max_task_priority::
   1291   1.1  mrg 
   1292   1.7  mrg _Reference_:
   1293   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.14
   1294   1.7  mrg 
   1295   1.7  mrg 
   1296   1.7  mrg File: libgomp.info,  Node: OMP_NESTED,  Next: OMP_NUM_THREADS,  Prev: OMP_MAX_TASK_PRIORITY,  Up: Environment Variables
   1297   1.7  mrg 
   1298  1.11  mrg 3.7 'OMP_NESTED' - Nested parallel regions
   1299   1.1  mrg ==========================================
   1300   1.1  mrg 
   1301   1.1  mrg _Description_:
   1302   1.1  mrg      Enable or disable nested parallel regions, i.e., whether team
   1303   1.5  mrg      members are allowed to create new teams.  The value of this
   1304  1.11  mrg      environment variable shall be 'TRUE' or 'FALSE'.  If undefined,
   1305   1.1  mrg      nested parallel regions are disabled by default.
   1306   1.1  mrg 
   1307   1.1  mrg _See also_:
   1308   1.1  mrg      *note omp_set_nested::
   1309   1.1  mrg 
   1310   1.1  mrg _Reference_:
   1311   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.6
   1312   1.1  mrg 
   1313   1.1  mrg 
   1314   1.5  mrg File: libgomp.info,  Node: OMP_NUM_THREADS,  Next: OMP_PROC_BIND,  Prev: OMP_NESTED,  Up: Environment Variables
   1315   1.1  mrg 
   1316  1.11  mrg 3.8 'OMP_NUM_THREADS' - Specifies the number of threads to use
   1317   1.1  mrg ==============================================================
   1318   1.1  mrg 
   1319   1.1  mrg _Description_:
   1320  1.11  mrg      Specifies the default number of threads to use in parallel regions.
   1321  1.11  mrg      The value of this variable shall be a comma-separated list of
   1322  1.11  mrg      positive integers; the value specified the number of threads to use
   1323  1.11  mrg      for the corresponding nested level.  If undefined one thread per
   1324  1.11  mrg      CPU is used.
   1325   1.1  mrg 
   1326   1.1  mrg _See also_:
   1327   1.1  mrg      *note omp_set_num_threads::
   1328   1.1  mrg 
   1329   1.1  mrg _Reference_:
   1330   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.2
   1331   1.1  mrg 
   1332   1.1  mrg 
   1333   1.5  mrg File: libgomp.info,  Node: OMP_PROC_BIND,  Next: OMP_PLACES,  Prev: OMP_NUM_THREADS,  Up: Environment Variables
   1334   1.5  mrg 
   1335  1.11  mrg 3.9 'OMP_PROC_BIND' - Whether theads may be moved between CPUs
   1336   1.5  mrg ==============================================================
   1337   1.5  mrg 
   1338   1.5  mrg _Description_:
   1339   1.5  mrg      Specifies whether threads may be moved between processors.  If set
   1340  1.11  mrg      to 'TRUE', OpenMP theads should not be moved; if set to 'FALSE'
   1341   1.5  mrg      they may be moved.  Alternatively, a comma separated list with the
   1342  1.11  mrg      values 'MASTER', 'CLOSE' and 'SPREAD' can be used to specify the
   1343   1.5  mrg      thread affinity policy for the corresponding nesting level.  With
   1344  1.11  mrg      'MASTER' the worker threads are in the same place partition as the
   1345  1.11  mrg      master thread.  With 'CLOSE' those are kept close to the master
   1346  1.11  mrg      thread in contiguous place partitions.  And with 'SPREAD' a sparse
   1347   1.5  mrg      distribution across the place partitions is used.
   1348   1.5  mrg 
   1349  1.11  mrg      When undefined, 'OMP_PROC_BIND' defaults to 'TRUE' when
   1350  1.11  mrg      'OMP_PLACES' or 'GOMP_CPU_AFFINITY' is set and 'FALSE' otherwise.
   1351   1.5  mrg 
   1352   1.5  mrg _See also_:
   1353   1.5  mrg      *note OMP_PLACES::, *note GOMP_CPU_AFFINITY::, *note
   1354   1.5  mrg      omp_get_proc_bind::
   1355   1.5  mrg 
   1356   1.5  mrg _Reference_:
   1357   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.4
   1358   1.5  mrg 
   1359   1.5  mrg 
   1360   1.5  mrg File: libgomp.info,  Node: OMP_PLACES,  Next: OMP_STACKSIZE,  Prev: OMP_PROC_BIND,  Up: Environment Variables
   1361   1.1  mrg 
   1362  1.11  mrg 3.10 'OMP_PLACES' - Specifies on which CPUs the theads should be placed
   1363   1.7  mrg =======================================================================
   1364   1.1  mrg 
   1365   1.1  mrg _Description_:
   1366  1.11  mrg      The thread placement can be either specified using an abstract name
   1367  1.11  mrg      or by an explicit list of the places.  The abstract names
   1368  1.11  mrg      'threads', 'cores' and 'sockets' can be optionally followed by a
   1369   1.5  mrg      positive number in parentheses, which denotes the how many places
   1370  1.11  mrg      shall be created.  With 'threads' each place corresponds to a
   1371  1.11  mrg      single hardware thread; 'cores' to a single core with the
   1372  1.11  mrg      corresponding number of hardware threads; and with 'sockets' the
   1373   1.5  mrg      place corresponds to a single socket.  The resulting placement can
   1374  1.11  mrg      be shown by setting the 'OMP_DISPLAY_ENV' environment variable.
   1375   1.5  mrg 
   1376   1.5  mrg      Alternatively, the placement can be specified explicitly as
   1377   1.5  mrg      comma-separated list of places.  A place is specified by set of
   1378   1.5  mrg      nonnegative numbers in curly braces, denoting the denoting the
   1379   1.5  mrg      hardware threads.  The hardware threads belonging to a place can
   1380   1.5  mrg      either be specified as comma-separated list of nonnegative thread
   1381   1.5  mrg      numbers or using an interval.  Multiple places can also be either
   1382   1.5  mrg      specified by a comma-separated list of places or by an interval.
   1383   1.5  mrg      To specify an interval, a colon followed by the count is placed
   1384   1.5  mrg      after after the hardware thread number or the place.  Optionally,
   1385   1.5  mrg      the length can be followed by a colon and the stride number -
   1386   1.5  mrg      otherwise a unit stride is assumed.  For instance, the following
   1387  1.11  mrg      specifies the same places list: '"{0,1,2}, {3,4,6}, {7,8,9},
   1388  1.11  mrg      {10,11,12}"'; '"{0:3}, {3:3}, {7:3}, {10:3}"'; and '"{0:2}:4:3"'.
   1389   1.5  mrg 
   1390  1.11  mrg      If 'OMP_PLACES' and 'GOMP_CPU_AFFINITY' are unset and
   1391  1.11  mrg      'OMP_PROC_BIND' is either unset or 'false', threads may be moved
   1392   1.5  mrg      between CPUs following no placement policy.
   1393   1.1  mrg 
   1394   1.1  mrg _See also_:
   1395   1.5  mrg      *note OMP_PROC_BIND::, *note GOMP_CPU_AFFINITY::, *note
   1396   1.5  mrg      omp_get_proc_bind::, *note OMP_DISPLAY_ENV::
   1397   1.1  mrg 
   1398   1.1  mrg _Reference_:
   1399   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.5
   1400   1.1  mrg 
   1401   1.1  mrg 
   1402   1.5  mrg File: libgomp.info,  Node: OMP_STACKSIZE,  Next: OMP_SCHEDULE,  Prev: OMP_PLACES,  Up: Environment Variables
   1403   1.1  mrg 
   1404  1.11  mrg 3.11 'OMP_STACKSIZE' - Set default thread stack size
   1405   1.5  mrg ====================================================
   1406   1.1  mrg 
   1407   1.1  mrg _Description_:
   1408   1.1  mrg      Set the default thread stack size in kilobytes, unless the number
   1409  1.11  mrg      is suffixed by 'B', 'K', 'M' or 'G', in which case the size is,
   1410  1.11  mrg      respectively, in bytes, kilobytes, megabytes or gigabytes.  This is
   1411  1.11  mrg      different from 'pthread_attr_setstacksize' which gets the number of
   1412  1.11  mrg      bytes as an argument.  If the stack size cannot be set due to
   1413  1.11  mrg      system constraints, an error is reported and the initial stack size
   1414  1.11  mrg      is left unchanged.  If undefined, the stack size is system
   1415  1.11  mrg      dependent.
   1416   1.1  mrg 
   1417   1.1  mrg _Reference_:
   1418   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.7
   1419   1.1  mrg 
   1420   1.1  mrg 
   1421   1.5  mrg File: libgomp.info,  Node: OMP_SCHEDULE,  Next: OMP_THREAD_LIMIT,  Prev: OMP_STACKSIZE,  Up: Environment Variables
   1422   1.1  mrg 
   1423  1.11  mrg 3.12 'OMP_SCHEDULE' - How threads are scheduled
   1424   1.5  mrg ===============================================
   1425   1.1  mrg 
   1426   1.1  mrg _Description_:
   1427  1.11  mrg      Allows to specify 'schedule type' and 'chunk size'.  The value of
   1428  1.11  mrg      the variable shall have the form: 'type[,chunk]' where 'type' is
   1429  1.11  mrg      one of 'static', 'dynamic', 'guided' or 'auto' The optional 'chunk'
   1430  1.11  mrg      size shall be a positive integer.  If undefined, dynamic scheduling
   1431  1.11  mrg      and a chunk size of 1 is used.
   1432   1.1  mrg 
   1433   1.1  mrg _See also_:
   1434   1.5  mrg      *note omp_set_schedule::
   1435   1.1  mrg 
   1436   1.1  mrg _Reference_:
   1437   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Sections
   1438   1.7  mrg      2.7.1.1 and 4.1
   1439   1.1  mrg 
   1440   1.1  mrg 
   1441   1.5  mrg File: libgomp.info,  Node: OMP_THREAD_LIMIT,  Next: OMP_WAIT_POLICY,  Prev: OMP_SCHEDULE,  Up: Environment Variables
   1442   1.1  mrg 
   1443  1.11  mrg 3.13 'OMP_THREAD_LIMIT' - Set the maximum number of threads
   1444   1.5  mrg ===========================================================
   1445   1.1  mrg 
   1446   1.1  mrg _Description_:
   1447   1.5  mrg      Specifies the number of threads to use for the whole program.  The
   1448   1.5  mrg      value of this variable shall be a positive integer.  If undefined,
   1449   1.5  mrg      the number of threads is not limited.
   1450   1.5  mrg 
   1451   1.5  mrg _See also_:
   1452   1.5  mrg      *note OMP_NUM_THREADS::, *note omp_get_thread_limit::
   1453   1.1  mrg 
   1454   1.1  mrg _Reference_:
   1455   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.10
   1456   1.3  mrg 
   1457   1.3  mrg 
   1458   1.5  mrg File: libgomp.info,  Node: OMP_WAIT_POLICY,  Next: GOMP_CPU_AFFINITY,  Prev: OMP_THREAD_LIMIT,  Up: Environment Variables
   1459   1.3  mrg 
   1460  1.11  mrg 3.14 'OMP_WAIT_POLICY' - How waiting threads are handled
   1461   1.5  mrg ========================================================
   1462   1.3  mrg 
   1463   1.3  mrg _Description_:
   1464   1.5  mrg      Specifies whether waiting threads should be active or passive.  If
   1465  1.11  mrg      the value is 'PASSIVE', waiting threads should not consume CPU
   1466  1.11  mrg      power while waiting; while the value is 'ACTIVE' specifies that
   1467   1.5  mrg      they should.  If undefined, threads wait actively for a short time
   1468   1.5  mrg      before waiting passively.
   1469   1.3  mrg 
   1470   1.3  mrg _See also_:
   1471   1.5  mrg      *note GOMP_SPINCOUNT::
   1472   1.3  mrg 
   1473   1.3  mrg _Reference_:
   1474   1.7  mrg      OpenMP specification v4.5 (http://www.openmp.org/), Section 4.8
   1475   1.1  mrg 
   1476   1.1  mrg 
   1477   1.5  mrg File: libgomp.info,  Node: GOMP_CPU_AFFINITY,  Next: GOMP_DEBUG,  Prev: OMP_WAIT_POLICY,  Up: Environment Variables
   1478   1.1  mrg 
   1479  1.11  mrg 3.15 'GOMP_CPU_AFFINITY' - Bind threads to specific CPUs
   1480   1.3  mrg ========================================================
   1481   1.1  mrg 
   1482   1.1  mrg _Description_:
   1483   1.5  mrg      Binds threads to specific CPUs.  The variable should contain a
   1484   1.5  mrg      space-separated or comma-separated list of CPUs.  This list may
   1485   1.3  mrg      contain different kinds of entries: either single CPU numbers in
   1486   1.3  mrg      any order, a range of CPUs (M-N) or a range with some stride
   1487  1.11  mrg      (M-N:S). CPU numbers are zero based.  For example,
   1488  1.11  mrg      'GOMP_CPU_AFFINITY="0 3 1-2 4-15:2"' will bind the initial thread
   1489   1.3  mrg      to CPU 0, the second to CPU 3, the third to CPU 1, the fourth to
   1490   1.3  mrg      CPU 2, the fifth to CPU 4, the sixth through tenth to CPUs 6, 8,
   1491   1.3  mrg      10, 12, and 14 respectively and then start assigning back from the
   1492  1.11  mrg      beginning of the list.  'GOMP_CPU_AFFINITY=0' binds all threads to
   1493   1.3  mrg      CPU 0.
   1494   1.1  mrg 
   1495   1.5  mrg      There is no libgomp library routine to determine whether a CPU
   1496   1.5  mrg      affinity specification is in effect.  As a workaround,
   1497  1.11  mrg      language-specific library functions, e.g., 'getenv' in C or
   1498  1.11  mrg      'GET_ENVIRONMENT_VARIABLE' in Fortran, may be used to query the
   1499  1.11  mrg      setting of the 'GOMP_CPU_AFFINITY' environment variable.  A defined
   1500  1.11  mrg      CPU affinity on startup cannot be changed or disabled during the
   1501  1.11  mrg      runtime of the application.
   1502  1.11  mrg 
   1503  1.11  mrg      If both 'GOMP_CPU_AFFINITY' and 'OMP_PROC_BIND' are set,
   1504  1.11  mrg      'OMP_PROC_BIND' has a higher precedence.  If neither has been set
   1505  1.11  mrg      and 'OMP_PROC_BIND' is unset, or when 'OMP_PROC_BIND' is set to
   1506  1.11  mrg      'FALSE', the host system will handle the assignment of threads to
   1507   1.5  mrg      CPUs.
   1508   1.1  mrg 
   1509   1.3  mrg _See also_:
   1510   1.5  mrg      *note OMP_PLACES::, *note OMP_PROC_BIND::
   1511   1.5  mrg 
   1512   1.5  mrg 
   1513   1.5  mrg File: libgomp.info,  Node: GOMP_DEBUG,  Next: GOMP_STACKSIZE,  Prev: GOMP_CPU_AFFINITY,  Up: Environment Variables
   1514   1.5  mrg 
   1515  1.11  mrg 3.16 'GOMP_DEBUG' - Enable debugging output
   1516   1.5  mrg ===========================================
   1517   1.5  mrg 
   1518   1.5  mrg _Description_:
   1519  1.11  mrg      Enable debugging output.  The variable should be set to '0'
   1520  1.11  mrg      (disabled, also the default if not set), or '1' (enabled).
   1521   1.5  mrg 
   1522   1.5  mrg      If enabled, some debugging output will be printed during execution.
   1523   1.5  mrg      This is currently not specified in more detail, and subject to
   1524   1.5  mrg      change.
   1525   1.3  mrg 
   1526   1.1  mrg 
   1527   1.5  mrg File: libgomp.info,  Node: GOMP_STACKSIZE,  Next: GOMP_SPINCOUNT,  Prev: GOMP_DEBUG,  Up: Environment Variables
   1528   1.1  mrg 
   1529  1.11  mrg 3.17 'GOMP_STACKSIZE' - Set default thread stack size
   1530   1.1  mrg =====================================================
   1531   1.1  mrg 
   1532   1.1  mrg _Description_:
   1533   1.5  mrg      Set the default thread stack size in kilobytes.  This is different
   1534  1.11  mrg      from 'pthread_attr_setstacksize' which gets the number of bytes as
   1535   1.5  mrg      an argument.  If the stack size cannot be set due to system
   1536   1.3  mrg      constraints, an error is reported and the initial stack size is
   1537   1.5  mrg      left unchanged.  If undefined, the stack size is system dependent.
   1538   1.1  mrg 
   1539   1.1  mrg _See also_:
   1540   1.3  mrg      *note OMP_STACKSIZE::
   1541   1.1  mrg 
   1542   1.1  mrg _Reference_:
   1543   1.1  mrg      GCC Patches Mailinglist
   1544   1.1  mrg      (http://gcc.gnu.org/ml/gcc-patches/2006-06/msg00493.html), GCC
   1545   1.1  mrg      Patches Mailinglist
   1546   1.1  mrg      (http://gcc.gnu.org/ml/gcc-patches/2006-06/msg00496.html)
   1547   1.1  mrg 
   1548   1.1  mrg 
   1549   1.7  mrg File: libgomp.info,  Node: GOMP_SPINCOUNT,  Next: GOMP_RTEMS_THREAD_POOLS,  Prev: GOMP_STACKSIZE,  Up: Environment Variables
   1550   1.5  mrg 
   1551  1.11  mrg 3.18 'GOMP_SPINCOUNT' - Set the busy-wait spin count
   1552   1.5  mrg ====================================================
   1553   1.5  mrg 
   1554   1.5  mrg _Description_:
   1555   1.5  mrg      Determines how long a threads waits actively with consuming CPU
   1556   1.5  mrg      power before waiting passively without consuming CPU power.  The
   1557  1.11  mrg      value may be either 'INFINITE', 'INFINITY' to always wait actively
   1558   1.5  mrg      or an integer which gives the number of spins of the busy-wait
   1559   1.5  mrg      loop.  The integer may optionally be followed by the following
   1560  1.11  mrg      suffixes acting as multiplication factors: 'k' (kilo, thousand),
   1561  1.11  mrg      'M' (mega, million), 'G' (giga, billion), or 'T' (tera, trillion).
   1562  1.11  mrg      If undefined, 0 is used when 'OMP_WAIT_POLICY' is 'PASSIVE',
   1563  1.11  mrg      300,000 is used when 'OMP_WAIT_POLICY' is undefined and 30 billion
   1564  1.11  mrg      is used when 'OMP_WAIT_POLICY' is 'ACTIVE'.  If there are more
   1565  1.11  mrg      OpenMP threads than available CPUs, 1000 and 100 spins are used for
   1566  1.11  mrg      'OMP_WAIT_POLICY' being 'ACTIVE' or undefined, respectively; unless
   1567  1.11  mrg      the 'GOMP_SPINCOUNT' is lower or 'OMP_WAIT_POLICY' is 'PASSIVE'.
   1568   1.5  mrg 
   1569   1.5  mrg _See also_:
   1570   1.5  mrg      *note OMP_WAIT_POLICY::
   1571   1.5  mrg 
   1572   1.5  mrg 
   1573   1.7  mrg File: libgomp.info,  Node: GOMP_RTEMS_THREAD_POOLS,  Prev: GOMP_SPINCOUNT,  Up: Environment Variables
   1574   1.7  mrg 
   1575  1.11  mrg 3.19 'GOMP_RTEMS_THREAD_POOLS' - Set the RTEMS specific thread pools
   1576   1.7  mrg ====================================================================
   1577   1.7  mrg 
   1578   1.7  mrg _Description_:
   1579   1.7  mrg      This environment variable is only used on the RTEMS real-time
   1580   1.7  mrg      operating system.  It determines the scheduler instance specific
   1581  1.11  mrg      thread pools.  The format for 'GOMP_RTEMS_THREAD_POOLS' is a list
   1582  1.11  mrg      of optional '<thread-pool-count>[$<priority>]@<scheduler-name>'
   1583  1.11  mrg      configurations separated by ':' where:
   1584  1.11  mrg         * '<thread-pool-count>' is the thread pool count for this
   1585   1.7  mrg           scheduler instance.
   1586  1.11  mrg         * '$<priority>' is an optional priority for the worker threads
   1587  1.11  mrg           of a thread pool according to 'pthread_setschedparam'.  In
   1588   1.7  mrg           case a priority value is omitted, then a worker thread will
   1589   1.7  mrg           inherit the priority of the OpenMP master thread that created
   1590   1.7  mrg           it.  The priority of the worker thread is not changed after
   1591   1.7  mrg           creation, even if a new OpenMP master thread using the worker
   1592   1.7  mrg           has a different priority.
   1593  1.11  mrg         * '@<scheduler-name>' is the scheduler instance name according
   1594   1.7  mrg           to the RTEMS application configuration.
   1595   1.7  mrg      In case no thread pool configuration is specified for a scheduler
   1596  1.11  mrg      instance, then each OpenMP master thread of this scheduler instance
   1597  1.11  mrg      will use its own dynamically allocated thread pool.  To limit the
   1598  1.11  mrg      worker thread count of the thread pools, each OpenMP master thread
   1599  1.11  mrg      must call 'omp_set_num_threads'.
   1600   1.7  mrg _Example_:
   1601  1.11  mrg      Lets suppose we have three scheduler instances 'IO', 'WRK0', and
   1602  1.11  mrg      'WRK1' with 'GOMP_RTEMS_THREAD_POOLS' set to '"1@WRK0:3$4@WRK1"'.
   1603   1.7  mrg      Then there are no thread pool restrictions for scheduler instance
   1604  1.11  mrg      'IO'.  In the scheduler instance 'WRK0' there is one thread pool
   1605   1.7  mrg      available.  Since no priority is specified for this scheduler
   1606   1.7  mrg      instance, the worker thread inherits the priority of the OpenMP
   1607  1.11  mrg      master thread that created it.  In the scheduler instance 'WRK1'
   1608  1.11  mrg      there are three thread pools available and their worker threads run
   1609  1.11  mrg      at priority four.
   1610   1.7  mrg 
   1611   1.7  mrg 
   1612   1.7  mrg File: libgomp.info,  Node: Enabling OpenACC,  Next: OpenACC Runtime Library Routines,  Prev: Environment Variables,  Up: Top
   1613   1.7  mrg 
   1614   1.7  mrg 4 Enabling OpenACC
   1615   1.7  mrg ******************
   1616   1.7  mrg 
   1617   1.7  mrg To activate the OpenACC extensions for C/C++ and Fortran, the
   1618  1.11  mrg compile-time flag '-fopenacc' must be specified.  This enables the
   1619  1.11  mrg OpenACC directive '#pragma acc' in C/C++ and '!$accp' directives in free
   1620  1.11  mrg form, 'c$acc', '*$acc' and '!$acc' directives in fixed form, '!$'
   1621  1.11  mrg conditional compilation sentinels in free form and 'c$', '*$' and '!$'
   1622   1.7  mrg sentinels in fixed form, for Fortran.  The flag also arranges for
   1623   1.7  mrg automatic linking of the OpenACC runtime library (*note OpenACC Runtime
   1624   1.7  mrg Library Routines::).
   1625   1.7  mrg 
   1626   1.7  mrg    A complete description of all OpenACC directives accepted may be
   1627  1.11  mrg found in the OpenACC (https://www.openacc.org) Application Programming
   1628   1.7  mrg Interface manual, version 2.0.
   1629   1.7  mrg 
   1630   1.7  mrg    Note that this is an experimental feature and subject to change in
   1631  1.11  mrg future versions of GCC. See <https://gcc.gnu.org/wiki/OpenACC> for more
   1632  1.11  mrg information.
   1633   1.7  mrg 
   1634   1.7  mrg 
   1635   1.7  mrg File: libgomp.info,  Node: OpenACC Runtime Library Routines,  Next: OpenACC Environment Variables,  Prev: Enabling OpenACC,  Up: Top
   1636   1.7  mrg 
   1637   1.7  mrg 5 OpenACC Runtime Library Routines
   1638   1.7  mrg **********************************
   1639   1.7  mrg 
   1640   1.7  mrg The runtime routines described here are defined by section 3 of the
   1641   1.7  mrg OpenACC specifications in version 2.0.  They have C linkage, and do not
   1642  1.11  mrg throw exceptions.  Generally, they are available only for the host, with
   1643  1.11  mrg the exception of 'acc_on_device', which is available for both the host
   1644  1.11  mrg and the acceleration device.
   1645   1.7  mrg 
   1646   1.7  mrg * Menu:
   1647   1.7  mrg 
   1648   1.7  mrg * acc_get_num_devices::         Get number of devices for the given device
   1649   1.7  mrg                                 type.
   1650   1.7  mrg * acc_set_device_type::         Set type of device accelerator to use.
   1651   1.7  mrg * acc_get_device_type::         Get type of device accelerator to be used.
   1652   1.7  mrg * acc_set_device_num::          Set device number to use.
   1653   1.7  mrg * acc_get_device_num::          Get device number to be used.
   1654   1.7  mrg * acc_async_test::              Tests for completion of a specific asynchronous
   1655   1.7  mrg                                 operation.
   1656   1.7  mrg * acc_async_test_all::          Tests for completion of all asychronous
   1657   1.7  mrg                                 operations.
   1658   1.7  mrg * acc_wait::                    Wait for completion of a specific asynchronous
   1659   1.7  mrg                                 operation.
   1660   1.7  mrg * acc_wait_all::                Waits for completion of all asyncrhonous
   1661   1.7  mrg                                 operations.
   1662   1.7  mrg * acc_wait_all_async::          Wait for completion of all asynchronous
   1663   1.7  mrg                                 operations.
   1664   1.7  mrg * acc_wait_async::              Wait for completion of asynchronous operations.
   1665   1.7  mrg * acc_init::                    Initialize runtime for a specific device type.
   1666   1.7  mrg * acc_shutdown::                Shuts down the runtime for a specific device
   1667   1.7  mrg                                 type.
   1668   1.7  mrg * acc_on_device::               Whether executing on a particular device
   1669   1.7  mrg * acc_malloc::                  Allocate device memory.
   1670   1.7  mrg * acc_free::                    Free device memory.
   1671   1.7  mrg * acc_copyin::                  Allocate device memory and copy host memory to
   1672   1.7  mrg                                 it.
   1673   1.7  mrg * acc_present_or_copyin::       If the data is not present on the device,
   1674   1.7  mrg                                 allocate device memory and copy from host
   1675   1.7  mrg                                 memory.
   1676   1.7  mrg * acc_create::                  Allocate device memory and map it to host
   1677   1.7  mrg                                 memory.
   1678   1.7  mrg * acc_present_or_create::       If the data is not present on the device,
   1679   1.7  mrg                                 allocate device memory and map it to host
   1680   1.7  mrg                                 memory.
   1681   1.7  mrg * acc_copyout::                 Copy device memory to host memory.
   1682   1.7  mrg * acc_delete::                  Free device memory.
   1683   1.7  mrg * acc_update_device::           Update device memory from mapped host memory.
   1684   1.7  mrg * acc_update_self::             Update host memory from mapped device memory.
   1685   1.7  mrg * acc_map_data::                Map previously allocated device memory to host
   1686   1.7  mrg                                 memory.
   1687   1.7  mrg * acc_unmap_data::              Unmap device memory from host memory.
   1688   1.7  mrg * acc_deviceptr::               Get device pointer associated with specific
   1689   1.7  mrg                                 host address.
   1690   1.7  mrg * acc_hostptr::                 Get host pointer associated with specific
   1691   1.7  mrg                                 device address.
   1692   1.7  mrg * acc_is_present::              Indiciate whether host variable / array is
   1693   1.7  mrg                                 present on device.
   1694   1.7  mrg * acc_memcpy_to_device::        Copy host memory to device memory.
   1695   1.7  mrg * acc_memcpy_from_device::      Copy device memory to host memory.
   1696   1.7  mrg 
   1697   1.7  mrg API routines for target platforms.
   1698   1.7  mrg 
   1699   1.7  mrg * acc_get_current_cuda_device:: Get CUDA device handle.
   1700   1.7  mrg * acc_get_current_cuda_context::Get CUDA context handle.
   1701   1.7  mrg * acc_get_cuda_stream::         Get CUDA stream handle.
   1702   1.7  mrg * acc_set_cuda_stream::         Set CUDA stream handle.
   1703   1.7  mrg 
   1704   1.7  mrg 
   1705   1.7  mrg File: libgomp.info,  Node: acc_get_num_devices,  Next: acc_set_device_type,  Up: OpenACC Runtime Library Routines
   1706   1.7  mrg 
   1707  1.11  mrg 5.1 'acc_get_num_devices' - Get number of devices for given device type
   1708   1.7  mrg =======================================================================
   1709   1.7  mrg 
   1710   1.7  mrg _Description_
   1711   1.7  mrg      This function returns a value indicating the number of devices
   1712   1.7  mrg      available for the device type specified in DEVICETYPE.
   1713   1.7  mrg 
   1714   1.7  mrg _C/C++_:
   1715  1.11  mrg      _Prototype_:   'int acc_get_num_devices(acc_device_t devicetype);'
   1716   1.7  mrg 
   1717   1.7  mrg _Fortran_:
   1718  1.11  mrg      _Interface_:   'integer function acc_get_num_devices(devicetype)'
   1719  1.11  mrg                     'integer(kind=acc_device_kind) devicetype'
   1720   1.7  mrg 
   1721   1.7  mrg _Reference_:
   1722  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   1723   1.7  mrg      3.2.1.
   1724   1.7  mrg 
   1725   1.7  mrg 
   1726   1.7  mrg File: libgomp.info,  Node: acc_set_device_type,  Next: acc_get_device_type,  Prev: acc_get_num_devices,  Up: OpenACC Runtime Library Routines
   1727   1.7  mrg 
   1728  1.11  mrg 5.2 'acc_set_device_type' - Set type of device accelerator to use.
   1729   1.7  mrg ==================================================================
   1730   1.7  mrg 
   1731   1.7  mrg _Description_
   1732   1.7  mrg      This function indicates to the runtime library which device typr,
   1733   1.7  mrg      specified in DEVICETYPE, to use when executing a parallel or
   1734   1.7  mrg      kernels region.
   1735   1.7  mrg 
   1736   1.7  mrg _C/C++_:
   1737  1.11  mrg      _Prototype_:   'acc_set_device_type(acc_device_t devicetype);'
   1738   1.7  mrg 
   1739   1.7  mrg _Fortran_:
   1740  1.11  mrg      _Interface_:   'subroutine acc_set_device_type(devicetype)'
   1741  1.11  mrg                     'integer(kind=acc_device_kind) devicetype'
   1742   1.7  mrg 
   1743   1.7  mrg _Reference_:
   1744  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   1745   1.7  mrg      3.2.2.
   1746   1.7  mrg 
   1747   1.7  mrg 
   1748   1.7  mrg File: libgomp.info,  Node: acc_get_device_type,  Next: acc_set_device_num,  Prev: acc_set_device_type,  Up: OpenACC Runtime Library Routines
   1749   1.7  mrg 
   1750  1.11  mrg 5.3 'acc_get_device_type' - Get type of device accelerator to be used.
   1751   1.7  mrg ======================================================================
   1752   1.7  mrg 
   1753   1.7  mrg _Description_
   1754   1.7  mrg      This function returns what device type will be used when executing
   1755   1.7  mrg      a parallel or kernels region.
   1756   1.7  mrg 
   1757   1.7  mrg _C/C++_:
   1758  1.11  mrg      _Prototype_:   'acc_device_t acc_get_device_type(void);'
   1759   1.7  mrg 
   1760   1.7  mrg _Fortran_:
   1761  1.11  mrg      _Interface_:   'function acc_get_device_type(void)'
   1762  1.11  mrg                     'integer(kind=acc_device_kind) acc_get_device_type'
   1763   1.7  mrg 
   1764   1.7  mrg _Reference_:
   1765  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   1766   1.7  mrg      3.2.3.
   1767   1.7  mrg 
   1768   1.7  mrg 
   1769   1.7  mrg File: libgomp.info,  Node: acc_set_device_num,  Next: acc_get_device_num,  Prev: acc_get_device_type,  Up: OpenACC Runtime Library Routines
   1770   1.7  mrg 
   1771  1.11  mrg 5.4 'acc_set_device_num' - Set device number to use.
   1772   1.7  mrg ====================================================
   1773   1.7  mrg 
   1774   1.7  mrg _Description_
   1775   1.7  mrg      This function will indicate to the runtime which device number,
   1776   1.7  mrg      specified by NUM, associated with the specifed device type
   1777   1.7  mrg      DEVICETYPE.
   1778   1.7  mrg 
   1779   1.7  mrg _C/C++_:
   1780  1.11  mrg      _Prototype_:   'acc_set_device_num(int num, acc_device_t devicetype);'
   1781   1.7  mrg 
   1782   1.7  mrg _Fortran_:
   1783  1.11  mrg      _Interface_:   'subroutine acc_set_device_num(devicenum, devicetype)'
   1784  1.11  mrg                     'integer devicenum'
   1785  1.11  mrg                     'integer(kind=acc_device_kind) devicetype'
   1786   1.7  mrg 
   1787   1.7  mrg _Reference_:
   1788  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   1789   1.7  mrg      3.2.4.
   1790   1.7  mrg 
   1791   1.7  mrg 
   1792   1.7  mrg File: libgomp.info,  Node: acc_get_device_num,  Next: acc_async_test,  Prev: acc_set_device_num,  Up: OpenACC Runtime Library Routines
   1793   1.7  mrg 
   1794  1.11  mrg 5.5 'acc_get_device_num' - Get device number to be used.
   1795   1.7  mrg ========================================================
   1796   1.7  mrg 
   1797   1.7  mrg _Description_
   1798   1.7  mrg      This function returns which device number associated with the
   1799   1.7  mrg      specified device type DEVICETYPE, will be used when executing a
   1800   1.7  mrg      parallel or kernels region.
   1801   1.7  mrg 
   1802   1.7  mrg _C/C++_:
   1803  1.11  mrg      _Prototype_:   'int acc_get_device_num(acc_device_t devicetype);'
   1804   1.7  mrg 
   1805   1.7  mrg _Fortran_:
   1806  1.11  mrg      _Interface_:   'function acc_get_device_num(devicetype)'
   1807  1.11  mrg                     'integer(kind=acc_device_kind) devicetype'
   1808  1.11  mrg                     'integer acc_get_device_num'
   1809   1.7  mrg 
   1810   1.7  mrg _Reference_:
   1811  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   1812   1.7  mrg      3.2.5.
   1813   1.7  mrg 
   1814   1.7  mrg 
   1815   1.7  mrg File: libgomp.info,  Node: acc_async_test,  Next: acc_async_test_all,  Prev: acc_get_device_num,  Up: OpenACC Runtime Library Routines
   1816   1.7  mrg 
   1817  1.11  mrg 5.6 'acc_async_test' - Test for completion of a specific asynchronous operation.
   1818   1.7  mrg ================================================================================
   1819   1.7  mrg 
   1820   1.7  mrg _Description_
   1821   1.7  mrg      This function tests for completion of the asynchrounous operation
   1822  1.11  mrg      specified in ARG.  In C/C++, a non-zero value will be returned to
   1823  1.11  mrg      indicate the specified asynchronous operation has completed.  While
   1824  1.11  mrg      Fortran will return a 'true'.  If the asynchrounous operation has
   1825  1.11  mrg      not completed, C/C++ returns a zero and Fortran returns a 'false'.
   1826   1.7  mrg 
   1827   1.7  mrg _C/C++_:
   1828  1.11  mrg      _Prototype_:   'int acc_async_test(int arg);'
   1829   1.7  mrg 
   1830   1.7  mrg _Fortran_:
   1831  1.11  mrg      _Interface_:   'function acc_async_test(arg)'
   1832  1.11  mrg                     'integer(kind=acc_handle_kind) arg'
   1833  1.11  mrg                     'logical acc_async_test'
   1834   1.7  mrg 
   1835   1.7  mrg _Reference_:
   1836  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   1837   1.7  mrg      3.2.6.
   1838   1.7  mrg 
   1839   1.7  mrg 
   1840   1.7  mrg File: libgomp.info,  Node: acc_async_test_all,  Next: acc_wait,  Prev: acc_async_test,  Up: OpenACC Runtime Library Routines
   1841   1.7  mrg 
   1842  1.11  mrg 5.7 'acc_async_test_all' - Tests for completion of all asynchronous operations.
   1843   1.7  mrg ===============================================================================
   1844   1.7  mrg 
   1845   1.7  mrg _Description_
   1846   1.7  mrg      This function tests for completion of all asynchrounous operations.
   1847   1.7  mrg      In C/C++, a non-zero value will be returned to indicate all
   1848  1.11  mrg      asynchronous operations have completed.  While Fortran will return
   1849  1.11  mrg      a 'true'.  If any asynchronous operation has not completed, C/C++
   1850  1.11  mrg      returns a zero and Fortran returns a 'false'.
   1851   1.7  mrg 
   1852   1.7  mrg _C/C++_:
   1853  1.11  mrg      _Prototype_:   'int acc_async_test_all(void);'
   1854   1.7  mrg 
   1855   1.7  mrg _Fortran_:
   1856  1.11  mrg      _Interface_:   'function acc_async_test()'
   1857  1.11  mrg                     'logical acc_get_device_num'
   1858   1.7  mrg 
   1859   1.7  mrg _Reference_:
   1860  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   1861   1.7  mrg      3.2.7.
   1862   1.7  mrg 
   1863   1.7  mrg 
   1864   1.7  mrg File: libgomp.info,  Node: acc_wait,  Next: acc_wait_all,  Prev: acc_async_test_all,  Up: OpenACC Runtime Library Routines
   1865   1.7  mrg 
   1866  1.11  mrg 5.8 'acc_wait' - Wait for completion of a specific asynchronous operation.
   1867   1.7  mrg ==========================================================================
   1868   1.7  mrg 
   1869   1.7  mrg _Description_
   1870   1.7  mrg      This function waits for completion of the asynchronous operation
   1871   1.7  mrg      specified in ARG.
   1872   1.7  mrg 
   1873   1.7  mrg _C/C++_:
   1874  1.11  mrg      _Prototype_:   'acc_wait(arg);'
   1875  1.11  mrg      _Prototype     'acc_async_wait(arg);'
   1876  1.11  mrg      (OpenACC 1.0
   1877  1.11  mrg      compatibility)_:
   1878   1.7  mrg 
   1879   1.7  mrg _Fortran_:
   1880  1.11  mrg      _Interface_:   'subroutine acc_wait(arg)'
   1881  1.11  mrg                     'integer(acc_handle_kind) arg'
   1882  1.11  mrg      _Interface     'subroutine acc_async_wait(arg)'
   1883  1.11  mrg      (OpenACC 1.0
   1884  1.11  mrg      compatibility)_:
   1885  1.11  mrg                     'integer(acc_handle_kind) arg'
   1886   1.7  mrg 
   1887   1.7  mrg _Reference_:
   1888  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   1889   1.7  mrg      3.2.8.
   1890   1.7  mrg 
   1891   1.7  mrg 
   1892   1.7  mrg File: libgomp.info,  Node: acc_wait_all,  Next: acc_wait_all_async,  Prev: acc_wait,  Up: OpenACC Runtime Library Routines
   1893   1.7  mrg 
   1894  1.11  mrg 5.9 'acc_wait_all' - Waits for completion of all asynchronous operations.
   1895   1.7  mrg =========================================================================
   1896   1.7  mrg 
   1897   1.7  mrg _Description_
   1898   1.7  mrg      This function waits for the completion of all asynchronous
   1899   1.7  mrg      operations.
   1900   1.7  mrg 
   1901   1.7  mrg _C/C++_:
   1902  1.11  mrg      _Prototype_:   'acc_wait_all(void);'
   1903  1.11  mrg      _Prototype     'acc_async_wait_all(void);'
   1904  1.11  mrg      (OpenACC 1.0
   1905  1.11  mrg      compatibility)_:
   1906   1.7  mrg 
   1907   1.7  mrg _Fortran_:
   1908  1.11  mrg      _Interface_:   'subroutine acc_wait_all()'
   1909  1.11  mrg      _Interface     'subroutine acc_async_wait_all()'
   1910  1.11  mrg      (OpenACC 1.0
   1911  1.11  mrg      compatibility)_:
   1912   1.7  mrg 
   1913   1.7  mrg _Reference_:
   1914  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   1915   1.7  mrg      3.2.10.
   1916   1.7  mrg 
   1917   1.7  mrg 
   1918   1.7  mrg File: libgomp.info,  Node: acc_wait_all_async,  Next: acc_wait_async,  Prev: acc_wait_all,  Up: OpenACC Runtime Library Routines
   1919   1.7  mrg 
   1920  1.11  mrg 5.10 'acc_wait_all_async' - Wait for completion of all asynchronous operations.
   1921   1.7  mrg ===============================================================================
   1922   1.7  mrg 
   1923   1.7  mrg _Description_
   1924   1.7  mrg      This function enqueues a wait operation on the queue ASYNC for any
   1925   1.7  mrg      and all asynchronous operations that have been previously enqueued
   1926   1.7  mrg      on any queue.
   1927   1.7  mrg 
   1928   1.7  mrg _C/C++_:
   1929  1.11  mrg      _Prototype_:   'acc_wait_all_async(int async);'
   1930   1.7  mrg 
   1931   1.7  mrg _Fortran_:
   1932  1.11  mrg      _Interface_:   'subroutine acc_wait_all_async(async)'
   1933  1.11  mrg                     'integer(acc_handle_kind) async'
   1934   1.7  mrg 
   1935   1.7  mrg _Reference_:
   1936  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   1937   1.7  mrg      3.2.11.
   1938   1.7  mrg 
   1939   1.7  mrg 
   1940   1.7  mrg File: libgomp.info,  Node: acc_wait_async,  Next: acc_init,  Prev: acc_wait_all_async,  Up: OpenACC Runtime Library Routines
   1941   1.7  mrg 
   1942  1.11  mrg 5.11 'acc_wait_async' - Wait for completion of asynchronous operations.
   1943   1.7  mrg =======================================================================
   1944   1.7  mrg 
   1945   1.7  mrg _Description_
   1946   1.7  mrg      This function enqueues a wait operation on queue ASYNC for any and
   1947   1.7  mrg      all asynchronous operations enqueued on queue ARG.
   1948   1.7  mrg 
   1949   1.7  mrg _C/C++_:
   1950  1.11  mrg      _Prototype_:   'acc_wait_async(int arg, int async);'
   1951   1.7  mrg 
   1952   1.7  mrg _Fortran_:
   1953  1.11  mrg      _Interface_:   'subroutine acc_wait_async(arg, async)'
   1954  1.11  mrg                     'integer(acc_handle_kind) arg, async'
   1955   1.7  mrg 
   1956   1.7  mrg _Reference_:
   1957  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   1958   1.7  mrg      3.2.9.
   1959   1.7  mrg 
   1960   1.7  mrg 
   1961   1.7  mrg File: libgomp.info,  Node: acc_init,  Next: acc_shutdown,  Prev: acc_wait_async,  Up: OpenACC Runtime Library Routines
   1962   1.7  mrg 
   1963  1.11  mrg 5.12 'acc_init' - Initialize runtime for a specific device type.
   1964   1.7  mrg ================================================================
   1965   1.7  mrg 
   1966   1.7  mrg _Description_
   1967  1.11  mrg      This function initializes the runtime for the device type specified
   1968  1.11  mrg      in DEVICETYPE.
   1969   1.7  mrg 
   1970   1.7  mrg _C/C++_:
   1971  1.11  mrg      _Prototype_:   'acc_init(acc_device_t devicetype);'
   1972   1.7  mrg 
   1973   1.7  mrg _Fortran_:
   1974  1.11  mrg      _Interface_:   'subroutine acc_init(devicetype)'
   1975  1.11  mrg                     'integer(acc_device_kind) devicetype'
   1976   1.7  mrg 
   1977   1.7  mrg _Reference_:
   1978  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   1979   1.7  mrg      3.2.12.
   1980   1.7  mrg 
   1981   1.7  mrg 
   1982   1.7  mrg File: libgomp.info,  Node: acc_shutdown,  Next: acc_on_device,  Prev: acc_init,  Up: OpenACC Runtime Library Routines
   1983   1.7  mrg 
   1984  1.11  mrg 5.13 'acc_shutdown' - Shuts down the runtime for a specific device type.
   1985   1.7  mrg ========================================================================
   1986   1.7  mrg 
   1987   1.7  mrg _Description_
   1988   1.7  mrg      This function shuts down the runtime for the device type specified
   1989   1.7  mrg      in DEVICETYPE.
   1990   1.7  mrg 
   1991   1.7  mrg _C/C++_:
   1992  1.11  mrg      _Prototype_:   'acc_shutdown(acc_device_t devicetype);'
   1993   1.7  mrg 
   1994   1.7  mrg _Fortran_:
   1995  1.11  mrg      _Interface_:   'subroutine acc_shutdown(devicetype)'
   1996  1.11  mrg                     'integer(acc_device_kind) devicetype'
   1997   1.7  mrg 
   1998   1.7  mrg _Reference_:
   1999  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2000   1.7  mrg      3.2.13.
   2001   1.7  mrg 
   2002   1.7  mrg 
   2003   1.7  mrg File: libgomp.info,  Node: acc_on_device,  Next: acc_malloc,  Prev: acc_shutdown,  Up: OpenACC Runtime Library Routines
   2004   1.7  mrg 
   2005  1.11  mrg 5.14 'acc_on_device' - Whether executing on a particular device
   2006   1.7  mrg ===============================================================
   2007   1.7  mrg 
   2008   1.7  mrg _Description_:
   2009   1.7  mrg      This function returns whether the program is executing on a
   2010  1.11  mrg      particular device specified in DEVICETYPE.  In C/C++ a non-zero
   2011   1.7  mrg      value is returned to indicate the device is execiting on the
   2012  1.11  mrg      specified device type.  In Fortran, 'true' will be returned.  If
   2013   1.7  mrg      the program is not executing on the specified device type C/C++
   2014  1.11  mrg      will return a zero, while Fortran will return 'false'.
   2015   1.7  mrg 
   2016   1.7  mrg _C/C++_:
   2017  1.11  mrg      _Prototype_:   'acc_on_device(acc_device_t devicetype);'
   2018   1.7  mrg 
   2019   1.7  mrg _Fortran_:
   2020  1.11  mrg      _Interface_:   'function acc_on_device(devicetype)'
   2021  1.11  mrg                     'integer(acc_device_kind) devicetype'
   2022  1.11  mrg                     'logical acc_on_device'
   2023   1.7  mrg 
   2024   1.7  mrg _Reference_:
   2025  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2026   1.7  mrg      3.2.14.
   2027   1.7  mrg 
   2028   1.7  mrg 
   2029   1.7  mrg File: libgomp.info,  Node: acc_malloc,  Next: acc_free,  Prev: acc_on_device,  Up: OpenACC Runtime Library Routines
   2030   1.7  mrg 
   2031  1.11  mrg 5.15 'acc_malloc' - Allocate device memory.
   2032   1.7  mrg ===========================================
   2033   1.7  mrg 
   2034   1.7  mrg _Description_
   2035  1.11  mrg      This function allocates LEN bytes of device memory.  It returns the
   2036   1.7  mrg      device address of the allocated memory.
   2037   1.7  mrg 
   2038   1.7  mrg _C/C++_:
   2039  1.11  mrg      _Prototype_:   'd_void* acc_malloc(size_t len);'
   2040   1.7  mrg 
   2041   1.7  mrg _Reference_:
   2042  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2043   1.7  mrg      3.2.15.
   2044   1.7  mrg 
   2045   1.7  mrg 
   2046   1.7  mrg File: libgomp.info,  Node: acc_free,  Next: acc_copyin,  Prev: acc_malloc,  Up: OpenACC Runtime Library Routines
   2047   1.7  mrg 
   2048  1.11  mrg 5.16 'acc_free' - Free device memory.
   2049   1.7  mrg =====================================
   2050   1.7  mrg 
   2051   1.7  mrg _Description_
   2052  1.11  mrg      Free previously allocated device memory at the device address 'a'.
   2053   1.7  mrg 
   2054   1.7  mrg _C/C++_:
   2055  1.11  mrg      _Prototype_:   'acc_free(d_void *a);'
   2056   1.7  mrg 
   2057   1.7  mrg _Reference_:
   2058  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2059   1.7  mrg      3.2.16.
   2060   1.7  mrg 
   2061   1.7  mrg 
   2062   1.7  mrg File: libgomp.info,  Node: acc_copyin,  Next: acc_present_or_copyin,  Prev: acc_free,  Up: OpenACC Runtime Library Routines
   2063   1.7  mrg 
   2064  1.11  mrg 5.17 'acc_copyin' - Allocate device memory and copy host memory to it.
   2065   1.7  mrg ======================================================================
   2066   1.7  mrg 
   2067   1.7  mrg _Description_
   2068   1.7  mrg      In C/C++, this function allocates LEN bytes of device memory and
   2069  1.11  mrg      maps it to the specified host address in A.  The device address of
   2070   1.7  mrg      the newly allocated device memory is returned.
   2071   1.7  mrg 
   2072  1.11  mrg      In Fortran, two (2) forms are supported.  In the first form, A
   2073  1.11  mrg      specifies a contiguous array section.  The second form A specifies
   2074  1.11  mrg      a variable or array element and LEN specifies the length in bytes.
   2075   1.7  mrg 
   2076   1.7  mrg _C/C++_:
   2077  1.11  mrg      _Prototype_:   'void *acc_copyin(h_void *a, size_t len);'
   2078   1.7  mrg 
   2079   1.7  mrg _Fortran_:
   2080  1.11  mrg      _Interface_:   'subroutine acc_copyin(a)'
   2081  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2082  1.11  mrg      _Interface_:   'subroutine acc_copyin(a, len)'
   2083  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2084  1.11  mrg                     'integer len'
   2085   1.7  mrg 
   2086   1.7  mrg _Reference_:
   2087  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2088   1.7  mrg      3.2.17.
   2089   1.7  mrg 
   2090   1.7  mrg 
   2091   1.7  mrg File: libgomp.info,  Node: acc_present_or_copyin,  Next: acc_create,  Prev: acc_copyin,  Up: OpenACC Runtime Library Routines
   2092   1.7  mrg 
   2093  1.11  mrg 5.18 'acc_present_or_copyin' - If the data is not present on the device, allocate device memory and copy from host memory.
   2094   1.7  mrg ==========================================================================================================================
   2095   1.7  mrg 
   2096   1.7  mrg _Description_
   2097   1.7  mrg      This function tests if the host data specifed by A and of length
   2098  1.11  mrg      LEN is present or not.  If it is not present, then device memory
   2099  1.11  mrg      will be allocated and the host memory copied.  The device address
   2100  1.11  mrg      of the newly allocated device memory is returned.
   2101   1.7  mrg 
   2102  1.11  mrg      In Fortran, two (2) forms are supported.  In the first form, A
   2103  1.11  mrg      specifies a contiguous array section.  The second form A specifies
   2104   1.7  mrg      a variable or array element and LEN specifies the length in bytes.
   2105   1.7  mrg 
   2106   1.7  mrg _C/C++_:
   2107  1.11  mrg      _Prototype_:   'void *acc_present_or_copyin(h_void *a, size_t len);'
   2108  1.11  mrg      _Prototype_:   'void *acc_pcopyin(h_void *a, size_t len);'
   2109   1.7  mrg 
   2110   1.7  mrg _Fortran_:
   2111  1.11  mrg      _Interface_:   'subroutine acc_present_or_copyin(a)'
   2112  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2113  1.11  mrg      _Interface_:   'subroutine acc_present_or_copyin(a, len)'
   2114  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2115  1.11  mrg                     'integer len'
   2116  1.11  mrg      _Interface_:   'subroutine acc_pcopyin(a)'
   2117  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2118  1.11  mrg      _Interface_:   'subroutine acc_pcopyin(a, len)'
   2119  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2120  1.11  mrg                     'integer len'
   2121   1.7  mrg 
   2122   1.7  mrg _Reference_:
   2123  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2124   1.7  mrg      3.2.18.
   2125   1.7  mrg 
   2126   1.7  mrg 
   2127   1.7  mrg File: libgomp.info,  Node: acc_create,  Next: acc_present_or_create,  Prev: acc_present_or_copyin,  Up: OpenACC Runtime Library Routines
   2128   1.7  mrg 
   2129  1.11  mrg 5.19 'acc_create' - Allocate device memory and map it to host memory.
   2130   1.7  mrg =====================================================================
   2131   1.7  mrg 
   2132   1.7  mrg _Description_
   2133   1.7  mrg      This function allocates device memory and maps it to host memory
   2134  1.11  mrg      specified by the host address A with a length of LEN bytes.  In
   2135   1.7  mrg      C/C++, the function returns the device address of the allocated
   2136   1.7  mrg      device memory.
   2137   1.7  mrg 
   2138  1.11  mrg      In Fortran, two (2) forms are supported.  In the first form, A
   2139  1.11  mrg      specifies a contiguous array section.  The second form A specifies
   2140   1.7  mrg      a variable or array element and LEN specifies the length in bytes.
   2141   1.7  mrg 
   2142   1.7  mrg _C/C++_:
   2143  1.11  mrg      _Prototype_:   'void *acc_create(h_void *a, size_t len);'
   2144   1.7  mrg 
   2145   1.7  mrg _Fortran_:
   2146  1.11  mrg      _Interface_:   'subroutine acc_create(a)'
   2147  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2148  1.11  mrg      _Interface_:   'subroutine acc_create(a, len)'
   2149  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2150  1.11  mrg                     'integer len'
   2151   1.7  mrg 
   2152   1.7  mrg _Reference_:
   2153  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2154   1.7  mrg      3.2.19.
   2155   1.7  mrg 
   2156   1.7  mrg 
   2157   1.7  mrg File: libgomp.info,  Node: acc_present_or_create,  Next: acc_copyout,  Prev: acc_create,  Up: OpenACC Runtime Library Routines
   2158   1.7  mrg 
   2159  1.11  mrg 5.20 'acc_present_or_create' - If the data is not present on the device, allocate device memory and map it to host memory.
   2160   1.7  mrg ==========================================================================================================================
   2161   1.1  mrg 
   2162   1.7  mrg _Description_
   2163   1.7  mrg      This function tests if the host data specifed by A and of length
   2164  1.11  mrg      LEN is present or not.  If it is not present, then device memory
   2165  1.11  mrg      will be allocated and mapped to host memory.  In C/C++, the device
   2166   1.7  mrg      address of the newly allocated device memory is returned.
   2167   1.7  mrg 
   2168  1.11  mrg      In Fortran, two (2) forms are supported.  In the first form, A
   2169  1.11  mrg      specifies a contiguous array section.  The second form A specifies
   2170   1.7  mrg      a variable or array element and LEN specifies the length in bytes.
   2171   1.7  mrg 
   2172   1.7  mrg _C/C++_:
   2173  1.11  mrg      _Prototype_:   'void *acc_present_or_create(h_void *a, size_t len)'
   2174  1.11  mrg      _Prototype_:   'void *acc_pcreate(h_void *a, size_t len)'
   2175   1.7  mrg 
   2176   1.7  mrg _Fortran_:
   2177  1.11  mrg      _Interface_:   'subroutine acc_present_or_create(a)'
   2178  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2179  1.11  mrg      _Interface_:   'subroutine acc_present_or_create(a, len)'
   2180  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2181  1.11  mrg                     'integer len'
   2182  1.11  mrg      _Interface_:   'subroutine acc_pcreate(a)'
   2183  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2184  1.11  mrg      _Interface_:   'subroutine acc_pcreate(a, len)'
   2185  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2186  1.11  mrg                     'integer len'
   2187   1.7  mrg 
   2188   1.7  mrg _Reference_:
   2189  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2190   1.7  mrg      3.2.20.
   2191   1.7  mrg 
   2192   1.7  mrg 
   2193   1.7  mrg File: libgomp.info,  Node: acc_copyout,  Next: acc_delete,  Prev: acc_present_or_create,  Up: OpenACC Runtime Library Routines
   2194   1.7  mrg 
   2195  1.11  mrg 5.21 'acc_copyout' - Copy device memory to host memory.
   2196   1.7  mrg =======================================================
   2197   1.7  mrg 
   2198   1.7  mrg _Description_
   2199   1.7  mrg      This function copies mapped device memory to host memory which is
   2200   1.7  mrg      specified by host address A for a length LEN bytes in C/C++.
   2201   1.7  mrg 
   2202  1.11  mrg      In Fortran, two (2) forms are supported.  In the first form, A
   2203  1.11  mrg      specifies a contiguous array section.  The second form A specifies
   2204   1.7  mrg      a variable or array element and LEN specifies the length in bytes.
   2205   1.7  mrg 
   2206   1.7  mrg _C/C++_:
   2207  1.11  mrg      _Prototype_:   'acc_copyout(h_void *a, size_t len);'
   2208   1.7  mrg 
   2209   1.7  mrg _Fortran_:
   2210  1.11  mrg      _Interface_:   'subroutine acc_copyout(a)'
   2211  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2212  1.11  mrg      _Interface_:   'subroutine acc_copyout(a, len)'
   2213  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2214  1.11  mrg                     'integer len'
   2215   1.7  mrg 
   2216   1.7  mrg _Reference_:
   2217  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2218   1.7  mrg      3.2.21.
   2219   1.7  mrg 
   2220   1.7  mrg 
   2221   1.7  mrg File: libgomp.info,  Node: acc_delete,  Next: acc_update_device,  Prev: acc_copyout,  Up: OpenACC Runtime Library Routines
   2222   1.7  mrg 
   2223  1.11  mrg 5.22 'acc_delete' - Free device memory.
   2224   1.7  mrg =======================================
   2225   1.7  mrg 
   2226   1.7  mrg _Description_
   2227   1.7  mrg      This function frees previously allocated device memory specified by
   2228   1.7  mrg      the device address A and the length of LEN bytes.
   2229   1.7  mrg 
   2230  1.11  mrg      In Fortran, two (2) forms are supported.  In the first form, A
   2231  1.11  mrg      specifies a contiguous array section.  The second form A specifies
   2232   1.7  mrg      a variable or array element and LEN specifies the length in bytes.
   2233   1.7  mrg 
   2234   1.7  mrg _C/C++_:
   2235  1.11  mrg      _Prototype_:   'acc_delete(h_void *a, size_t len);'
   2236   1.7  mrg 
   2237   1.7  mrg _Fortran_:
   2238  1.11  mrg      _Interface_:   'subroutine acc_delete(a)'
   2239  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2240  1.11  mrg      _Interface_:   'subroutine acc_delete(a, len)'
   2241  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2242  1.11  mrg                     'integer len'
   2243   1.7  mrg 
   2244   1.7  mrg _Reference_:
   2245  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2246   1.7  mrg      3.2.22.
   2247   1.7  mrg 
   2248   1.7  mrg 
   2249   1.7  mrg File: libgomp.info,  Node: acc_update_device,  Next: acc_update_self,  Prev: acc_delete,  Up: OpenACC Runtime Library Routines
   2250   1.7  mrg 
   2251  1.11  mrg 5.23 'acc_update_device' - Update device memory from mapped host memory.
   2252   1.7  mrg ========================================================================
   2253   1.7  mrg 
   2254   1.7  mrg _Description_
   2255   1.7  mrg      This function updates the device copy from the previously mapped
   2256   1.7  mrg      host memory.  The host memory is specified with the host address A
   2257   1.7  mrg      and a length of LEN bytes.
   2258   1.7  mrg 
   2259  1.11  mrg      In Fortran, two (2) forms are supported.  In the first form, A
   2260  1.11  mrg      specifies a contiguous array section.  The second form A specifies
   2261   1.7  mrg      a variable or array element and LEN specifies the length in bytes.
   2262   1.7  mrg 
   2263   1.7  mrg _C/C++_:
   2264  1.11  mrg      _Prototype_:   'acc_update_device(h_void *a, size_t len);'
   2265   1.7  mrg 
   2266   1.7  mrg _Fortran_:
   2267  1.11  mrg      _Interface_:   'subroutine acc_update_device(a)'
   2268  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2269  1.11  mrg      _Interface_:   'subroutine acc_update_device(a, len)'
   2270  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2271  1.11  mrg                     'integer len'
   2272   1.7  mrg 
   2273   1.7  mrg _Reference_:
   2274  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2275   1.7  mrg      3.2.23.
   2276   1.7  mrg 
   2277   1.7  mrg 
   2278   1.7  mrg File: libgomp.info,  Node: acc_update_self,  Next: acc_map_data,  Prev: acc_update_device,  Up: OpenACC Runtime Library Routines
   2279   1.7  mrg 
   2280  1.11  mrg 5.24 'acc_update_self' - Update host memory from mapped device memory.
   2281   1.7  mrg ======================================================================
   2282   1.7  mrg 
   2283   1.7  mrg _Description_
   2284   1.7  mrg      This function updates the host copy from the previously mapped
   2285   1.7  mrg      device memory.  The host memory is specified with the host address
   2286   1.7  mrg      A and a length of LEN bytes.
   2287   1.7  mrg 
   2288  1.11  mrg      In Fortran, two (2) forms are supported.  In the first form, A
   2289  1.11  mrg      specifies a contiguous array section.  The second form A specifies
   2290   1.7  mrg      a variable or array element and LEN specifies the length in bytes.
   2291   1.7  mrg 
   2292   1.7  mrg _C/C++_:
   2293  1.11  mrg      _Prototype_:   'acc_update_self(h_void *a, size_t len);'
   2294   1.7  mrg 
   2295   1.7  mrg _Fortran_:
   2296  1.11  mrg      _Interface_:   'subroutine acc_update_self(a)'
   2297  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2298  1.11  mrg      _Interface_:   'subroutine acc_update_self(a, len)'
   2299  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2300  1.11  mrg                     'integer len'
   2301   1.7  mrg 
   2302   1.7  mrg _Reference_:
   2303  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2304   1.7  mrg      3.2.24.
   2305   1.7  mrg 
   2306   1.7  mrg 
   2307   1.7  mrg File: libgomp.info,  Node: acc_map_data,  Next: acc_unmap_data,  Prev: acc_update_self,  Up: OpenACC Runtime Library Routines
   2308   1.7  mrg 
   2309  1.11  mrg 5.25 'acc_map_data' - Map previously allocated device memory to host memory.
   2310   1.7  mrg ============================================================================
   2311   1.7  mrg 
   2312   1.7  mrg _Description_
   2313   1.7  mrg      This function maps previously allocated device and host memory.
   2314  1.11  mrg      The device memory is specified with the device address D.  The host
   2315   1.7  mrg      memory is specified with the host address H and a length of LEN.
   2316   1.7  mrg 
   2317   1.7  mrg _C/C++_:
   2318  1.11  mrg      _Prototype_:   'acc_map_data(h_void *h, d_void *d, size_t len);'
   2319   1.7  mrg 
   2320   1.7  mrg _Reference_:
   2321  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2322   1.7  mrg      3.2.25.
   2323   1.7  mrg 
   2324   1.7  mrg 
   2325   1.7  mrg File: libgomp.info,  Node: acc_unmap_data,  Next: acc_deviceptr,  Prev: acc_map_data,  Up: OpenACC Runtime Library Routines
   2326   1.7  mrg 
   2327  1.11  mrg 5.26 'acc_unmap_data' - Unmap device memory from host memory.
   2328   1.7  mrg =============================================================
   2329   1.7  mrg 
   2330   1.7  mrg _Description_
   2331  1.11  mrg      This function unmaps previously mapped device and host memory.  The
   2332   1.7  mrg      latter specified by H.
   2333   1.7  mrg 
   2334   1.7  mrg _C/C++_:
   2335  1.11  mrg      _Prototype_:   'acc_unmap_data(h_void *h);'
   2336   1.7  mrg 
   2337   1.7  mrg _Reference_:
   2338  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2339   1.7  mrg      3.2.26.
   2340   1.7  mrg 
   2341   1.7  mrg 
   2342   1.7  mrg File: libgomp.info,  Node: acc_deviceptr,  Next: acc_hostptr,  Prev: acc_unmap_data,  Up: OpenACC Runtime Library Routines
   2343   1.7  mrg 
   2344  1.11  mrg 5.27 'acc_deviceptr' - Get device pointer associated with specific host address.
   2345   1.7  mrg ================================================================================
   2346   1.7  mrg 
   2347   1.7  mrg _Description_
   2348   1.7  mrg      This function returns the device address that has been mapped to
   2349   1.7  mrg      the host address specified by H.
   2350   1.7  mrg 
   2351   1.7  mrg _C/C++_:
   2352  1.11  mrg      _Prototype_:   'void *acc_deviceptr(h_void *h);'
   2353   1.7  mrg 
   2354   1.7  mrg _Reference_:
   2355  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2356   1.7  mrg      3.2.27.
   2357   1.7  mrg 
   2358   1.7  mrg 
   2359   1.7  mrg File: libgomp.info,  Node: acc_hostptr,  Next: acc_is_present,  Prev: acc_deviceptr,  Up: OpenACC Runtime Library Routines
   2360   1.7  mrg 
   2361  1.11  mrg 5.28 'acc_hostptr' - Get host pointer associated with specific device address.
   2362   1.7  mrg ==============================================================================
   2363   1.7  mrg 
   2364   1.7  mrg _Description_
   2365   1.7  mrg      This function returns the host address that has been mapped to the
   2366   1.7  mrg      device address specified by D.
   2367   1.7  mrg 
   2368   1.7  mrg _C/C++_:
   2369  1.11  mrg      _Prototype_:   'void *acc_hostptr(d_void *d);'
   2370   1.7  mrg 
   2371   1.7  mrg _Reference_:
   2372  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2373   1.7  mrg      3.2.28.
   2374   1.7  mrg 
   2375   1.7  mrg 
   2376   1.7  mrg File: libgomp.info,  Node: acc_is_present,  Next: acc_memcpy_to_device,  Prev: acc_hostptr,  Up: OpenACC Runtime Library Routines
   2377   1.7  mrg 
   2378  1.11  mrg 5.29 'acc_is_present' - Indicate whether host variable / array is present on device.
   2379   1.7  mrg ====================================================================================
   2380   1.7  mrg 
   2381   1.7  mrg _Description_
   2382  1.11  mrg      This function indicates whether the specified host address in A and
   2383  1.11  mrg      a length of LEN bytes is present on the device.  In C/C++, a
   2384   1.7  mrg      non-zero value is returned to indicate the presence of the mapped
   2385  1.11  mrg      memory on the device.  A zero is returned to indicate the memory is
   2386   1.7  mrg      not mapped on the device.
   2387   1.7  mrg 
   2388  1.11  mrg      In Fortran, two (2) forms are supported.  In the first form, A
   2389  1.11  mrg      specifies a contiguous array section.  The second form A specifies
   2390   1.7  mrg      a variable or array element and LEN specifies the length in bytes.
   2391  1.11  mrg      If the host memory is mapped to device memory, then a 'true' is
   2392  1.11  mrg      returned.  Otherwise, a 'false' is return to indicate the mapped
   2393   1.7  mrg      memory is not present.
   2394   1.7  mrg 
   2395   1.7  mrg _C/C++_:
   2396  1.11  mrg      _Prototype_:   'int acc_is_present(h_void *a, size_t len);'
   2397   1.7  mrg 
   2398   1.7  mrg _Fortran_:
   2399  1.11  mrg      _Interface_:   'function acc_is_present(a)'
   2400  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2401  1.11  mrg                     'logical acc_is_present'
   2402  1.11  mrg      _Interface_:   'function acc_is_present(a, len)'
   2403  1.11  mrg                     'type, dimension(:[,:]...) :: a'
   2404  1.11  mrg                     'integer len'
   2405  1.11  mrg                     'logical acc_is_present'
   2406   1.7  mrg 
   2407   1.7  mrg _Reference_:
   2408  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2409   1.7  mrg      3.2.29.
   2410   1.7  mrg 
   2411   1.7  mrg 
   2412   1.7  mrg File: libgomp.info,  Node: acc_memcpy_to_device,  Next: acc_memcpy_from_device,  Prev: acc_is_present,  Up: OpenACC Runtime Library Routines
   2413   1.7  mrg 
   2414  1.11  mrg 5.30 'acc_memcpy_to_device' - Copy host memory to device memory.
   2415   1.7  mrg ================================================================
   2416   1.7  mrg 
   2417   1.7  mrg _Description_
   2418   1.7  mrg      This function copies host memory specified by host address of SRC
   2419   1.7  mrg      to device memory specified by the device address DEST for a length
   2420   1.7  mrg      of BYTES bytes.
   2421   1.7  mrg 
   2422   1.7  mrg _C/C++_:
   2423  1.11  mrg      _Prototype_:   'acc_memcpy_to_device(d_void *dest, h_void *src, size_t
   2424  1.11  mrg                     bytes);'
   2425   1.7  mrg 
   2426   1.7  mrg _Reference_:
   2427  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2428   1.7  mrg      3.2.30.
   2429   1.7  mrg 
   2430   1.7  mrg 
   2431   1.7  mrg File: libgomp.info,  Node: acc_memcpy_from_device,  Next: acc_get_current_cuda_device,  Prev: acc_memcpy_to_device,  Up: OpenACC Runtime Library Routines
   2432   1.7  mrg 
   2433  1.11  mrg 5.31 'acc_memcpy_from_device' - Copy device memory to host memory.
   2434   1.7  mrg ==================================================================
   2435   1.7  mrg 
   2436   1.7  mrg _Description_
   2437   1.7  mrg      This function copies host memory specified by host address of SRC
   2438   1.7  mrg      from device memory specified by the device address DEST for a
   2439   1.7  mrg      length of BYTES bytes.
   2440   1.7  mrg 
   2441   1.7  mrg _C/C++_:
   2442  1.11  mrg      _Prototype_:   'acc_memcpy_from_device(d_void *dest, h_void *src,
   2443  1.11  mrg                     size_t bytes);'
   2444   1.7  mrg 
   2445   1.7  mrg _Reference_:
   2446  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2447   1.7  mrg      3.2.31.
   2448   1.7  mrg 
   2449   1.7  mrg 
   2450   1.7  mrg File: libgomp.info,  Node: acc_get_current_cuda_device,  Next: acc_get_current_cuda_context,  Prev: acc_memcpy_from_device,  Up: OpenACC Runtime Library Routines
   2451   1.7  mrg 
   2452  1.11  mrg 5.32 'acc_get_current_cuda_device' - Get CUDA device handle.
   2453   1.7  mrg ============================================================
   2454   1.7  mrg 
   2455   1.7  mrg _Description_
   2456  1.11  mrg      This function returns the CUDA device handle.  This handle is the
   2457   1.7  mrg      same as used by the CUDA Runtime or Driver API's.
   2458   1.7  mrg 
   2459   1.7  mrg _C/C++_:
   2460  1.11  mrg      _Prototype_:   'void *acc_get_current_cuda_device(void);'
   2461   1.7  mrg 
   2462   1.7  mrg _Reference_:
   2463  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2464   1.7  mrg      A.2.1.1.
   2465   1.7  mrg 
   2466   1.7  mrg 
   2467   1.7  mrg File: libgomp.info,  Node: acc_get_current_cuda_context,  Next: acc_get_cuda_stream,  Prev: acc_get_current_cuda_device,  Up: OpenACC Runtime Library Routines
   2468   1.7  mrg 
   2469  1.11  mrg 5.33 'acc_get_current_cuda_context' - Get CUDA context handle.
   2470   1.7  mrg ==============================================================
   2471   1.7  mrg 
   2472   1.7  mrg _Description_
   2473  1.11  mrg      This function returns the CUDA context handle.  This handle is the
   2474   1.7  mrg      same as used by the CUDA Runtime or Driver API's.
   2475   1.7  mrg 
   2476   1.7  mrg _C/C++_:
   2477  1.11  mrg      _Prototype_:   'acc_get_current_cuda_context(void);'
   2478   1.7  mrg 
   2479   1.7  mrg _Reference_:
   2480  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2481   1.7  mrg      A.2.1.2.
   2482   1.7  mrg 
   2483   1.7  mrg 
   2484   1.7  mrg File: libgomp.info,  Node: acc_get_cuda_stream,  Next: acc_set_cuda_stream,  Prev: acc_get_current_cuda_context,  Up: OpenACC Runtime Library Routines
   2485   1.7  mrg 
   2486  1.11  mrg 5.34 'acc_get_cuda_stream' - Get CUDA stream handle.
   2487   1.7  mrg ====================================================
   2488   1.7  mrg 
   2489   1.7  mrg _Description_
   2490  1.11  mrg      This function returns the CUDA stream handle.  This handle is the
   2491   1.7  mrg      same as used by the CUDA Runtime or Driver API's.
   2492   1.7  mrg 
   2493   1.7  mrg _C/C++_:
   2494  1.11  mrg      _Prototype_:   'acc_get_cuda_stream(void);'
   2495   1.7  mrg 
   2496   1.7  mrg _Reference_:
   2497  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2498   1.7  mrg      A.2.1.3.
   2499   1.7  mrg 
   2500   1.7  mrg 
   2501   1.7  mrg File: libgomp.info,  Node: acc_set_cuda_stream,  Prev: acc_get_cuda_stream,  Up: OpenACC Runtime Library Routines
   2502   1.7  mrg 
   2503  1.11  mrg 5.35 'acc_set_cuda_stream' - Set CUDA stream handle.
   2504   1.7  mrg ====================================================
   2505   1.7  mrg 
   2506   1.7  mrg _Description_
   2507   1.7  mrg      This function associates the stream handle specified by STREAM with
   2508   1.7  mrg      the asynchronous value specified by ASYNC.
   2509   1.7  mrg 
   2510   1.7  mrg _C/C++_:
   2511  1.11  mrg      _Prototype_:   'acc_set_cuda_stream(int async void *stream);'
   2512   1.7  mrg 
   2513   1.7  mrg _Reference_:
   2514  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section
   2515   1.7  mrg      A.2.1.4.
   2516   1.7  mrg 
   2517   1.7  mrg 
   2518   1.7  mrg File: libgomp.info,  Node: OpenACC Environment Variables,  Next: CUDA Streams Usage,  Prev: OpenACC Runtime Library Routines,  Up: Top
   2519   1.7  mrg 
   2520   1.7  mrg 6 OpenACC Environment Variables
   2521   1.7  mrg *******************************
   2522   1.7  mrg 
   2523  1.11  mrg The variables 'ACC_DEVICE_TYPE' and 'ACC_DEVICE_NUM' are defined by
   2524   1.7  mrg section 4 of the OpenACC specification in version 2.0.  The variable
   2525  1.11  mrg 'GCC_ACC_NOTIFY' is used for diagnostic purposes.
   2526   1.7  mrg 
   2527   1.7  mrg * Menu:
   2528   1.7  mrg 
   2529   1.7  mrg * ACC_DEVICE_TYPE::
   2530   1.7  mrg * ACC_DEVICE_NUM::
   2531   1.7  mrg * GCC_ACC_NOTIFY::
   2532   1.7  mrg 
   2533   1.7  mrg 
   2534   1.7  mrg File: libgomp.info,  Node: ACC_DEVICE_TYPE,  Next: ACC_DEVICE_NUM,  Up: OpenACC Environment Variables
   2535   1.7  mrg 
   2536  1.11  mrg 6.1 'ACC_DEVICE_TYPE'
   2537   1.7  mrg =====================
   2538   1.7  mrg 
   2539   1.7  mrg _Reference_:
   2540  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section 4.1.
   2541   1.7  mrg 
   2542   1.7  mrg 
   2543   1.7  mrg File: libgomp.info,  Node: ACC_DEVICE_NUM,  Next: GCC_ACC_NOTIFY,  Prev: ACC_DEVICE_TYPE,  Up: OpenACC Environment Variables
   2544   1.7  mrg 
   2545  1.11  mrg 6.2 'ACC_DEVICE_NUM'
   2546   1.7  mrg ====================
   2547   1.7  mrg 
   2548   1.7  mrg _Reference_:
   2549  1.11  mrg      OpenACC specification v2.0 (https://www.openacc.org), section 4.2.
   2550   1.7  mrg 
   2551   1.7  mrg 
   2552   1.7  mrg File: libgomp.info,  Node: GCC_ACC_NOTIFY,  Prev: ACC_DEVICE_NUM,  Up: OpenACC Environment Variables
   2553   1.7  mrg 
   2554  1.11  mrg 6.3 'GCC_ACC_NOTIFY'
   2555   1.7  mrg ====================
   2556   1.7  mrg 
   2557   1.7  mrg _Description_:
   2558   1.7  mrg      Print debug information pertaining to the accelerator.
   2559   1.7  mrg 
   2560   1.7  mrg 
   2561   1.7  mrg File: libgomp.info,  Node: CUDA Streams Usage,  Next: OpenACC Library Interoperability,  Prev: OpenACC Environment Variables,  Up: Top
   2562   1.7  mrg 
   2563   1.7  mrg 7 CUDA Streams Usage
   2564   1.7  mrg ********************
   2565   1.7  mrg 
   2566  1.11  mrg This applies to the 'nvptx' plugin only.
   2567   1.7  mrg 
   2568   1.7  mrg    The library provides elements that perform asynchronous movement of
   2569   1.7  mrg data and asynchronous operation of computing constructs.  This
   2570   1.7  mrg asynchronous functionality is implemented by making use of CUDA
   2571   1.7  mrg streams(1).
   2572   1.7  mrg 
   2573   1.7  mrg    The primary means by that the asychronous functionality is accessed
   2574   1.7  mrg is through the use of those OpenACC directives which make use of the
   2575  1.11  mrg 'async' and 'wait' clauses.  When the 'async' clause is first used with
   2576  1.11  mrg a directive, it creates a CUDA stream.  If an 'async-argument' is used
   2577  1.11  mrg with the 'async' clause, then the stream is associated with the
   2578  1.11  mrg specified 'async-argument'.
   2579   1.7  mrg 
   2580   1.7  mrg    Following the creation of an association between a CUDA stream and
   2581  1.11  mrg the 'async-argument' of an 'async' clause, both the 'wait' clause and
   2582  1.11  mrg the 'wait' directive can be used.  When either the clause or directive
   2583   1.7  mrg is used after stream creation, it creates a rendezvous point whereby
   2584   1.7  mrg execution waits until all operations associated with the
   2585  1.11  mrg 'async-argument', that is, stream, have completed.
   2586   1.7  mrg 
   2587   1.7  mrg    Normally, the management of the streams that are created as a result
   2588  1.11  mrg of using the 'async' clause, is done without any intervention by the
   2589  1.11  mrg caller.  This implies the association between the 'async-argument' and
   2590   1.7  mrg the CUDA stream will be maintained for the lifetime of the program.
   2591   1.7  mrg However, this association can be changed through the use of the library
   2592  1.11  mrg function 'acc_set_cuda_stream'.  When the function 'acc_set_cuda_stream'
   2593  1.11  mrg is called, the CUDA stream that was originally associated with the
   2594  1.11  mrg 'async' clause will be destroyed.  Caution should be taken when changing
   2595  1.11  mrg the association as subsequent references to the 'async-argument' refer
   2596  1.11  mrg to a different CUDA stream.
   2597   1.7  mrg 
   2598   1.7  mrg    ---------- Footnotes ----------
   2599   1.7  mrg 
   2600   1.7  mrg    (1) See "Stream Management" in "CUDA Driver API", TRM-06703-001,
   2601   1.7  mrg Version 5.5, for additional information
   2602   1.7  mrg 
   2603   1.7  mrg 
   2604   1.7  mrg File: libgomp.info,  Node: OpenACC Library Interoperability,  Next: The libgomp ABI,  Prev: CUDA Streams Usage,  Up: Top
   2605   1.7  mrg 
   2606   1.7  mrg 8 OpenACC Library Interoperability
   2607   1.7  mrg **********************************
   2608   1.7  mrg 
   2609   1.7  mrg 8.1 Introduction
   2610   1.7  mrg ================
   2611   1.7  mrg 
   2612   1.7  mrg The OpenACC library uses the CUDA Driver API, and may interact with
   2613  1.11  mrg programs that use the Runtime library directly, or another library based
   2614  1.11  mrg on the Runtime library, e.g., CUBLAS(1). This chapter describes the use
   2615  1.11  mrg cases and what changes are required in order to use both the OpenACC
   2616  1.11  mrg library and the CUBLAS and Runtime libraries within a program.
   2617   1.7  mrg 
   2618   1.7  mrg 8.2 First invocation: NVIDIA CUBLAS library API
   2619   1.7  mrg ===============================================
   2620   1.7  mrg 
   2621   1.7  mrg In this first use case (see below), a function in the CUBLAS library is
   2622  1.11  mrg called prior to any of the functions in the OpenACC library.  More
   2623  1.11  mrg specifically, the function 'cublasCreate()'.
   2624   1.7  mrg 
   2625   1.7  mrg    When invoked, the function initializes the library and allocates the
   2626   1.7  mrg hardware resources on the host and the device on behalf of the caller.
   2627   1.7  mrg Once the initialization and allocation has completed, a handle is
   2628  1.11  mrg returned to the caller.  The OpenACC library also requires
   2629  1.11  mrg initialization and allocation of hardware resources.  Since the CUBLAS
   2630  1.11  mrg library has already allocated the hardware resources for the device, all
   2631  1.11  mrg that is left to do is to initialize the OpenACC library and acquire the
   2632  1.11  mrg hardware resources on the host.
   2633   1.7  mrg 
   2634   1.7  mrg    Prior to calling the OpenACC function that initializes the library
   2635  1.11  mrg and allocate the host hardware resources, you need to acquire the device
   2636  1.11  mrg number that was allocated during the call to 'cublasCreate()'.  The
   2637  1.11  mrg invoking of the runtime library function 'cudaGetDevice()' accomplishes
   2638  1.11  mrg this.  Once acquired, the device number is passed along with the device
   2639  1.11  mrg type as parameters to the OpenACC library function
   2640  1.11  mrg 'acc_set_device_num()'.
   2641  1.11  mrg 
   2642  1.11  mrg    Once the call to 'acc_set_device_num()' has completed, the OpenACC
   2643  1.11  mrg library uses the context that was created during the call to
   2644  1.11  mrg 'cublasCreate()'.  In other words, both libraries will be sharing the
   2645   1.7  mrg same context.
   2646   1.7  mrg 
   2647   1.7  mrg          /* Create the handle */
   2648   1.7  mrg          s = cublasCreate(&h);
   2649   1.7  mrg          if (s != CUBLAS_STATUS_SUCCESS)
   2650   1.7  mrg          {
   2651   1.7  mrg              fprintf(stderr, "cublasCreate failed %d\n", s);
   2652   1.7  mrg              exit(EXIT_FAILURE);
   2653   1.7  mrg          }
   2654   1.7  mrg 
   2655   1.7  mrg          /* Get the device number */
   2656   1.7  mrg          e = cudaGetDevice(&dev);
   2657   1.7  mrg          if (e != cudaSuccess)
   2658   1.7  mrg          {
   2659   1.7  mrg              fprintf(stderr, "cudaGetDevice failed %d\n", e);
   2660   1.7  mrg              exit(EXIT_FAILURE);
   2661   1.7  mrg          }
   2662   1.7  mrg 
   2663   1.7  mrg          /* Initialize OpenACC library and use device 'dev' */
   2664   1.7  mrg          acc_set_device_num(dev, acc_device_nvidia);
   2665  1.11  mrg 
   2666  1.11  mrg                               Use Case 1
   2667   1.7  mrg 
   2668   1.7  mrg 8.3 First invocation: OpenACC library API
   2669   1.7  mrg =========================================
   2670   1.7  mrg 
   2671   1.7  mrg In this second use case (see below), a function in the OpenACC library
   2672  1.11  mrg is called prior to any of the functions in the CUBLAS library.  More
   2673  1.11  mrg specificially, the function 'acc_set_device_num()'.
   2674   1.7  mrg 
   2675  1.11  mrg    In the use case presented here, the function 'acc_set_device_num()'
   2676   1.7  mrg is used to both initialize the OpenACC library and allocate the hardware
   2677  1.11  mrg resources on the host and the device.  In the call to the function, the
   2678  1.11  mrg call parameters specify which device to use and what device type to use,
   2679  1.11  mrg i.e., 'acc_device_nvidia'.  It should be noted that this is but one
   2680   1.7  mrg method to initialize the OpenACC library and allocate the appropriate
   2681  1.11  mrg hardware resources.  Other methods are available through the use of
   2682   1.7  mrg environment variables and these will be discussed in the next section.
   2683   1.7  mrg 
   2684  1.11  mrg    Once the call to 'acc_set_device_num()' has completed, other OpenACC
   2685   1.7  mrg functions can be called as seen with multiple calls being made to
   2686  1.11  mrg 'acc_copyin()'.  In addition, calls can be made to functions in the
   2687  1.11  mrg CUBLAS library.  In the use case a call to 'cublasCreate()' is made
   2688  1.11  mrg subsequent to the calls to 'acc_copyin()'.  As seen in the previous use
   2689  1.11  mrg case, a call to 'cublasCreate()' initializes the CUBLAS library and
   2690   1.7  mrg allocates the hardware resources on the host and the device.  However,
   2691  1.11  mrg since the device has already been allocated, 'cublasCreate()' will only
   2692   1.7  mrg initialize the CUBLAS library and allocate the appropriate hardware
   2693  1.11  mrg resources on the host.  The context that was created as part of the
   2694   1.7  mrg OpenACC initialization is shared with the CUBLAS library, similarly to
   2695   1.7  mrg the first use case.
   2696   1.7  mrg 
   2697   1.7  mrg          dev = 0;
   2698   1.7  mrg 
   2699   1.7  mrg          acc_set_device_num(dev, acc_device_nvidia);
   2700   1.7  mrg 
   2701   1.7  mrg          /* Copy the first set to the device */
   2702   1.7  mrg          d_X = acc_copyin(&h_X[0], N * sizeof (float));
   2703   1.7  mrg          if (d_X == NULL)
   2704   1.7  mrg          {
   2705   1.7  mrg              fprintf(stderr, "copyin error h_X\n");
   2706   1.7  mrg              exit(EXIT_FAILURE);
   2707   1.7  mrg          }
   2708   1.7  mrg 
   2709   1.7  mrg          /* Copy the second set to the device */
   2710   1.7  mrg          d_Y = acc_copyin(&h_Y1[0], N * sizeof (float));
   2711   1.7  mrg          if (d_Y == NULL)
   2712   1.7  mrg          {
   2713   1.7  mrg              fprintf(stderr, "copyin error h_Y1\n");
   2714   1.7  mrg              exit(EXIT_FAILURE);
   2715   1.7  mrg          }
   2716   1.7  mrg 
   2717   1.7  mrg          /* Create the handle */
   2718   1.7  mrg          s = cublasCreate(&h);
   2719   1.7  mrg          if (s != CUBLAS_STATUS_SUCCESS)
   2720   1.7  mrg          {
   2721   1.7  mrg              fprintf(stderr, "cublasCreate failed %d\n", s);
   2722   1.7  mrg              exit(EXIT_FAILURE);
   2723   1.7  mrg          }
   2724   1.7  mrg 
   2725   1.7  mrg          /* Perform saxpy using CUBLAS library function */
   2726   1.7  mrg          s = cublasSaxpy(h, N, &alpha, d_X, 1, d_Y, 1);
   2727   1.7  mrg          if (s != CUBLAS_STATUS_SUCCESS)
   2728   1.7  mrg          {
   2729   1.7  mrg              fprintf(stderr, "cublasSaxpy failed %d\n", s);
   2730   1.7  mrg              exit(EXIT_FAILURE);
   2731   1.7  mrg          }
   2732   1.7  mrg 
   2733   1.7  mrg          /* Copy the results from the device */
   2734   1.7  mrg          acc_memcpy_from_device(&h_Y1[0], d_Y, N * sizeof (float));
   2735  1.11  mrg 
   2736  1.11  mrg                               Use Case 2
   2737   1.7  mrg 
   2738   1.7  mrg 8.4 OpenACC library and environment variables
   2739   1.7  mrg =============================================
   2740   1.7  mrg 
   2741   1.7  mrg There are two environment variables associated with the OpenACC library
   2742   1.7  mrg that may be used to control the device type and device number:
   2743  1.11  mrg 'ACC_DEVICE_TYPE' and 'ACC_DEVICE_NUM', respecively.  These two
   2744   1.7  mrg environement variables can be used as an alternative to calling
   2745  1.11  mrg 'acc_set_device_num()'.  As seen in the second use case, the device type
   2746  1.11  mrg and device number were specified using 'acc_set_device_num()'.  If
   2747   1.7  mrg however, the aforementioned environment variables were set, then the
   2748  1.11  mrg call to 'acc_set_device_num()' would not be required.
   2749   1.7  mrg 
   2750  1.11  mrg    The use of the environment variables is only relevant when an OpenACC
   2751  1.11  mrg function is called prior to a call to 'cudaCreate()'.  If 'cudaCreate()'
   2752  1.11  mrg is called prior to a call to an OpenACC function, then you must call
   2753  1.11  mrg 'acc_set_device_num()'(2)
   2754   1.7  mrg 
   2755   1.7  mrg    ---------- Footnotes ----------
   2756   1.7  mrg 
   2757   1.7  mrg    (1) See section 2.26, "Interactions with the CUDA Driver API" in
   2758   1.7  mrg "CUDA Runtime API", Version 5.5, and section 2.27, "VDPAU
   2759  1.11  mrg Interoperability", in "CUDA Driver API", TRM-06703-001, Version 5.5, for
   2760  1.11  mrg additional information on library interoperability.
   2761   1.7  mrg 
   2762  1.11  mrg    (2) More complete information about 'ACC_DEVICE_TYPE' and
   2763  1.11  mrg 'ACC_DEVICE_NUM' can be found in sections 4.1 and 4.2 of the OpenACC
   2764  1.11  mrg (https://www.openacc.org) Application Programming Interface, Version
   2765   1.7  mrg 2.0.
   2766   1.7  mrg 
   2767   1.7  mrg 
   2768   1.7  mrg File: libgomp.info,  Node: The libgomp ABI,  Next: Reporting Bugs,  Prev: OpenACC Library Interoperability,  Up: Top
   2769   1.7  mrg 
   2770   1.7  mrg 9 The libgomp ABI
   2771   1.1  mrg *****************
   2772   1.1  mrg 
   2773  1.11  mrg The following sections present notes on the external ABI as presented by
   2774  1.11  mrg libgomp.  Only maintainers should need them.
   2775   1.1  mrg 
   2776   1.1  mrg * Menu:
   2777   1.1  mrg 
   2778   1.1  mrg * Implementing MASTER construct::
   2779   1.1  mrg * Implementing CRITICAL construct::
   2780   1.1  mrg * Implementing ATOMIC construct::
   2781   1.1  mrg * Implementing FLUSH construct::
   2782   1.1  mrg * Implementing BARRIER construct::
   2783   1.1  mrg * Implementing THREADPRIVATE construct::
   2784   1.1  mrg * Implementing PRIVATE clause::
   2785   1.1  mrg * Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses::
   2786   1.1  mrg * Implementing REDUCTION clause::
   2787   1.1  mrg * Implementing PARALLEL construct::
   2788   1.1  mrg * Implementing FOR construct::
   2789   1.1  mrg * Implementing ORDERED construct::
   2790   1.1  mrg * Implementing SECTIONS construct::
   2791   1.1  mrg * Implementing SINGLE construct::
   2792   1.7  mrg * Implementing OpenACC's PARALLEL construct::
   2793   1.1  mrg 
   2794   1.1  mrg 
   2795   1.1  mrg File: libgomp.info,  Node: Implementing MASTER construct,  Next: Implementing CRITICAL construct,  Up: The libgomp ABI
   2796   1.1  mrg 
   2797   1.7  mrg 9.1 Implementing MASTER construct
   2798   1.1  mrg =================================
   2799   1.1  mrg 
   2800   1.1  mrg      if (omp_get_thread_num () == 0)
   2801   1.1  mrg        block
   2802   1.1  mrg 
   2803   1.1  mrg    Alternately, we generate two copies of the parallel subfunction and
   2804   1.3  mrg only include this in the version run by the master thread.  Surely this
   2805   1.3  mrg is not worthwhile though...
   2806   1.1  mrg 
   2807   1.1  mrg 
   2808   1.1  mrg File: libgomp.info,  Node: Implementing CRITICAL construct,  Next: Implementing ATOMIC construct,  Prev: Implementing MASTER construct,  Up: The libgomp ABI
   2809   1.1  mrg 
   2810   1.7  mrg 9.2 Implementing CRITICAL construct
   2811   1.1  mrg ===================================
   2812   1.1  mrg 
   2813   1.1  mrg Without a specified name,
   2814   1.1  mrg 
   2815   1.1  mrg        void GOMP_critical_start (void);
   2816   1.1  mrg        void GOMP_critical_end (void);
   2817   1.1  mrg 
   2818   1.1  mrg    so that we don't get COPY relocations from libgomp to the main
   2819   1.1  mrg application.
   2820   1.1  mrg 
   2821   1.1  mrg    With a specified name, use omp_set_lock and omp_unset_lock with name
   2822   1.1  mrg being transformed into a variable declared like
   2823   1.1  mrg 
   2824   1.1  mrg        omp_lock_t gomp_critical_user_<name> __attribute__((common))
   2825   1.1  mrg 
   2826   1.1  mrg    Ideally the ABI would specify that all zero is a valid unlocked
   2827   1.3  mrg state, and so we wouldn't need to initialize this at startup.
   2828   1.1  mrg 
   2829   1.1  mrg 
   2830   1.1  mrg File: libgomp.info,  Node: Implementing ATOMIC construct,  Next: Implementing FLUSH construct,  Prev: Implementing CRITICAL construct,  Up: The libgomp ABI
   2831   1.1  mrg 
   2832   1.7  mrg 9.3 Implementing ATOMIC construct
   2833   1.1  mrg =================================
   2834   1.1  mrg 
   2835  1.11  mrg The target should implement the '__sync' builtins.
   2836   1.1  mrg 
   2837   1.1  mrg    Failing that we could add
   2838   1.1  mrg 
   2839   1.1  mrg        void GOMP_atomic_enter (void)
   2840   1.1  mrg        void GOMP_atomic_exit (void)
   2841   1.1  mrg 
   2842   1.1  mrg    which reuses the regular lock code, but with yet another lock object
   2843   1.1  mrg private to the library.
   2844   1.1  mrg 
   2845   1.1  mrg 
   2846   1.1  mrg File: libgomp.info,  Node: Implementing FLUSH construct,  Next: Implementing BARRIER construct,  Prev: Implementing ATOMIC construct,  Up: The libgomp ABI
   2847   1.1  mrg 
   2848   1.7  mrg 9.4 Implementing FLUSH construct
   2849   1.1  mrg ================================
   2850   1.1  mrg 
   2851  1.11  mrg Expands to the '__sync_synchronize' builtin.
   2852   1.1  mrg 
   2853   1.1  mrg 
   2854   1.1  mrg File: libgomp.info,  Node: Implementing BARRIER construct,  Next: Implementing THREADPRIVATE construct,  Prev: Implementing FLUSH construct,  Up: The libgomp ABI
   2855   1.1  mrg 
   2856   1.7  mrg 9.5 Implementing BARRIER construct
   2857   1.1  mrg ==================================
   2858   1.1  mrg 
   2859   1.1  mrg        void GOMP_barrier (void)
   2860   1.1  mrg 
   2861   1.1  mrg 
   2862   1.1  mrg File: libgomp.info,  Node: Implementing THREADPRIVATE construct,  Next: Implementing PRIVATE clause,  Prev: Implementing BARRIER construct,  Up: The libgomp ABI
   2863   1.1  mrg 
   2864   1.7  mrg 9.6 Implementing THREADPRIVATE construct
   2865   1.1  mrg ========================================
   2866   1.1  mrg 
   2867  1.11  mrg In _most_ cases we can map this directly to '__thread'.  Except that OMP
   2868  1.11  mrg allows constructors for C++ objects.  We can either refuse to support
   2869  1.11  mrg this (how often is it used?)  or we can implement something akin to
   2870  1.11  mrg .ctors.
   2871   1.1  mrg 
   2872   1.1  mrg    Even more ideally, this ctor feature is handled by extensions to the
   2873   1.1  mrg main pthreads library.  Failing that, we can have a set of entry points
   2874   1.1  mrg to register ctor functions to be called.
   2875   1.1  mrg 
   2876   1.1  mrg 
   2877   1.1  mrg File: libgomp.info,  Node: Implementing PRIVATE clause,  Next: Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses,  Prev: Implementing THREADPRIVATE construct,  Up: The libgomp ABI
   2878   1.1  mrg 
   2879   1.7  mrg 9.7 Implementing PRIVATE clause
   2880   1.1  mrg ===============================
   2881   1.1  mrg 
   2882   1.1  mrg In association with a PARALLEL, or within the lexical extent of a
   2883   1.1  mrg PARALLEL block, the variable becomes a local variable in the parallel
   2884   1.1  mrg subfunction.
   2885   1.1  mrg 
   2886   1.1  mrg    In association with FOR or SECTIONS blocks, create a new automatic
   2887   1.1  mrg variable within the current function.  This preserves the semantic of
   2888   1.1  mrg new variable creation.
   2889   1.1  mrg 
   2890   1.1  mrg 
   2891   1.1  mrg File: libgomp.info,  Node: Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses,  Next: Implementing REDUCTION clause,  Prev: Implementing PRIVATE clause,  Up: The libgomp ABI
   2892   1.1  mrg 
   2893   1.7  mrg 9.8 Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses
   2894   1.1  mrg ========================================================================
   2895   1.1  mrg 
   2896   1.3  mrg This seems simple enough for PARALLEL blocks.  Create a private struct
   2897   1.3  mrg for communicating between the parent and subfunction.  In the parent,
   2898   1.3  mrg copy in values for scalar and "small" structs; copy in addresses for
   2899   1.3  mrg others TREE_ADDRESSABLE types.  In the subfunction, copy the value into
   2900   1.3  mrg the local variable.
   2901   1.1  mrg 
   2902  1.11  mrg    It is not clear what to do with bare FOR or SECTION blocks.  The only
   2903  1.11  mrg thing I can figure is that we do something like:
   2904   1.1  mrg 
   2905   1.1  mrg      #pragma omp for firstprivate(x) lastprivate(y)
   2906   1.1  mrg      for (int i = 0; i < n; ++i)
   2907   1.1  mrg        body;
   2908   1.1  mrg 
   2909   1.1  mrg    which becomes
   2910   1.1  mrg 
   2911   1.1  mrg      {
   2912   1.1  mrg        int x = x, y;
   2913   1.1  mrg 
   2914   1.1  mrg        // for stuff
   2915   1.1  mrg 
   2916   1.1  mrg        if (i == n)
   2917   1.1  mrg          y = y;
   2918   1.1  mrg      }
   2919   1.1  mrg 
   2920   1.1  mrg    where the "x=x" and "y=y" assignments actually have different uids
   2921  1.11  mrg for the two variables, i.e.  not something you could write directly in
   2922  1.11  mrg C. Presumably this only makes sense if the "outer" x and y are global
   2923   1.1  mrg variables.
   2924   1.1  mrg 
   2925   1.1  mrg    COPYPRIVATE would work the same way, except the structure broadcast
   2926   1.1  mrg would have to happen via SINGLE machinery instead.
   2927   1.1  mrg 
   2928   1.1  mrg 
   2929   1.1  mrg File: libgomp.info,  Node: Implementing REDUCTION clause,  Next: Implementing PARALLEL construct,  Prev: Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses,  Up: The libgomp ABI
   2930   1.1  mrg 
   2931   1.7  mrg 9.9 Implementing REDUCTION clause
   2932   1.1  mrg =================================
   2933   1.1  mrg 
   2934   1.1  mrg The private struct mentioned in the previous section should have a
   2935  1.11  mrg pointer to an array of the type of the variable, indexed by the thread's
   2936  1.11  mrg TEAM_ID.  The thread stores its final value into the array, and after
   2937  1.11  mrg the barrier, the master thread iterates over the array to collect the
   2938  1.11  mrg values.
   2939   1.1  mrg 
   2940   1.1  mrg 
   2941   1.1  mrg File: libgomp.info,  Node: Implementing PARALLEL construct,  Next: Implementing FOR construct,  Prev: Implementing REDUCTION clause,  Up: The libgomp ABI
   2942   1.1  mrg 
   2943   1.7  mrg 9.10 Implementing PARALLEL construct
   2944   1.1  mrg ====================================
   2945   1.1  mrg 
   2946   1.1  mrg        #pragma omp parallel
   2947   1.1  mrg        {
   2948   1.1  mrg          body;
   2949   1.1  mrg        }
   2950   1.1  mrg 
   2951   1.1  mrg    becomes
   2952   1.1  mrg 
   2953   1.1  mrg        void subfunction (void *data)
   2954   1.1  mrg        {
   2955   1.1  mrg          use data;
   2956   1.1  mrg          body;
   2957   1.1  mrg        }
   2958   1.1  mrg 
   2959   1.1  mrg        setup data;
   2960   1.1  mrg        GOMP_parallel_start (subfunction, &data, num_threads);
   2961   1.1  mrg        subfunction (&data);
   2962   1.1  mrg        GOMP_parallel_end ();
   2963   1.1  mrg 
   2964   1.1  mrg        void GOMP_parallel_start (void (*fn)(void *), void *data, unsigned num_threads)
   2965   1.1  mrg 
   2966   1.1  mrg    The FN argument is the subfunction to be run in parallel.
   2967   1.1  mrg 
   2968   1.1  mrg    The DATA argument is a pointer to a structure used to communicate
   2969   1.1  mrg data in and out of the subfunction, as discussed above with respect to
   2970   1.1  mrg FIRSTPRIVATE et al.
   2971   1.1  mrg 
   2972   1.1  mrg    The NUM_THREADS argument is 1 if an IF clause is present and false,
   2973   1.1  mrg or the value of the NUM_THREADS clause, if present, or 0.
   2974   1.1  mrg 
   2975  1.11  mrg    The function needs to create the appropriate number of threads and/or
   2976  1.11  mrg launch them from the dock.  It needs to create the team structure and
   2977  1.11  mrg assign team ids.
   2978   1.1  mrg 
   2979   1.1  mrg        void GOMP_parallel_end (void)
   2980   1.1  mrg 
   2981   1.1  mrg    Tears down the team and returns us to the previous
   2982  1.11  mrg 'omp_in_parallel()' state.
   2983   1.1  mrg 
   2984   1.1  mrg 
   2985   1.1  mrg File: libgomp.info,  Node: Implementing FOR construct,  Next: Implementing ORDERED construct,  Prev: Implementing PARALLEL construct,  Up: The libgomp ABI
   2986   1.1  mrg 
   2987   1.7  mrg 9.11 Implementing FOR construct
   2988   1.1  mrg ===============================
   2989   1.1  mrg 
   2990   1.1  mrg        #pragma omp parallel for
   2991   1.1  mrg        for (i = lb; i <= ub; i++)
   2992   1.1  mrg          body;
   2993   1.1  mrg 
   2994   1.1  mrg    becomes
   2995   1.1  mrg 
   2996   1.1  mrg        void subfunction (void *data)
   2997   1.1  mrg        {
   2998   1.1  mrg          long _s0, _e0;
   2999   1.1  mrg          while (GOMP_loop_static_next (&_s0, &_e0))
   3000   1.1  mrg          {
   3001   1.1  mrg            long _e1 = _e0, i;
   3002   1.1  mrg            for (i = _s0; i < _e1; i++)
   3003   1.1  mrg              body;
   3004   1.1  mrg          }
   3005   1.1  mrg          GOMP_loop_end_nowait ();
   3006   1.1  mrg        }
   3007   1.1  mrg 
   3008   1.1  mrg        GOMP_parallel_loop_static (subfunction, NULL, 0, lb, ub+1, 1, 0);
   3009   1.1  mrg        subfunction (NULL);
   3010   1.1  mrg        GOMP_parallel_end ();
   3011   1.1  mrg 
   3012   1.1  mrg        #pragma omp for schedule(runtime)
   3013   1.1  mrg        for (i = 0; i < n; i++)
   3014   1.1  mrg          body;
   3015   1.1  mrg 
   3016   1.1  mrg    becomes
   3017   1.1  mrg 
   3018   1.1  mrg        {
   3019   1.1  mrg          long i, _s0, _e0;
   3020   1.1  mrg          if (GOMP_loop_runtime_start (0, n, 1, &_s0, &_e0))
   3021   1.1  mrg            do {
   3022   1.1  mrg              long _e1 = _e0;
   3023   1.1  mrg              for (i = _s0, i < _e0; i++)
   3024   1.1  mrg                body;
   3025   1.1  mrg            } while (GOMP_loop_runtime_next (&_s0, _&e0));
   3026   1.1  mrg          GOMP_loop_end ();
   3027   1.1  mrg        }
   3028   1.1  mrg 
   3029   1.3  mrg    Note that while it looks like there is trickiness to propagating a
   3030   1.1  mrg non-constant STEP, there isn't really.  We're explicitly allowed to
   3031   1.1  mrg evaluate it as many times as we want, and any variables involved should
   3032   1.1  mrg automatically be handled as PRIVATE or SHARED like any other variables.
   3033   1.1  mrg So the expression should remain evaluable in the subfunction.  We can
   3034  1.11  mrg also pull it into a local variable if we like, but since its supposed to
   3035  1.11  mrg remain unchanged, we can also not if we like.
   3036   1.1  mrg 
   3037  1.11  mrg    If we have SCHEDULE(STATIC), and no ORDERED, then we ought to be able
   3038  1.11  mrg to get away with no work-sharing context at all, since we can simply
   3039  1.11  mrg perform the arithmetic directly in each thread to divide up the
   3040  1.11  mrg iterations.  Which would mean that we wouldn't need to call any of these
   3041  1.11  mrg routines.
   3042   1.1  mrg 
   3043   1.1  mrg    There are separate routines for handling loops with an ORDERED
   3044   1.1  mrg clause.  Bookkeeping for that is non-trivial...
   3045   1.1  mrg 
   3046   1.1  mrg 
   3047   1.1  mrg File: libgomp.info,  Node: Implementing ORDERED construct,  Next: Implementing SECTIONS construct,  Prev: Implementing FOR construct,  Up: The libgomp ABI
   3048   1.1  mrg 
   3049   1.7  mrg 9.12 Implementing ORDERED construct
   3050   1.1  mrg ===================================
   3051   1.1  mrg 
   3052   1.1  mrg        void GOMP_ordered_start (void)
   3053   1.1  mrg        void GOMP_ordered_end (void)
   3054   1.1  mrg 
   3055   1.1  mrg 
   3056   1.1  mrg File: libgomp.info,  Node: Implementing SECTIONS construct,  Next: Implementing SINGLE construct,  Prev: Implementing ORDERED construct,  Up: The libgomp ABI
   3057   1.1  mrg 
   3058   1.7  mrg 9.13 Implementing SECTIONS construct
   3059   1.1  mrg ====================================
   3060   1.1  mrg 
   3061   1.1  mrg A block as
   3062   1.1  mrg 
   3063   1.1  mrg        #pragma omp sections
   3064   1.1  mrg        {
   3065   1.1  mrg          #pragma omp section
   3066   1.1  mrg          stmt1;
   3067   1.1  mrg          #pragma omp section
   3068   1.1  mrg          stmt2;
   3069   1.1  mrg          #pragma omp section
   3070   1.1  mrg          stmt3;
   3071   1.1  mrg        }
   3072   1.1  mrg 
   3073   1.1  mrg    becomes
   3074   1.1  mrg 
   3075   1.1  mrg        for (i = GOMP_sections_start (3); i != 0; i = GOMP_sections_next ())
   3076   1.1  mrg          switch (i)
   3077   1.1  mrg            {
   3078   1.1  mrg            case 1:
   3079   1.1  mrg              stmt1;
   3080   1.1  mrg              break;
   3081   1.1  mrg            case 2:
   3082   1.1  mrg              stmt2;
   3083   1.1  mrg              break;
   3084   1.1  mrg            case 3:
   3085   1.1  mrg              stmt3;
   3086   1.1  mrg              break;
   3087   1.1  mrg            }
   3088   1.1  mrg        GOMP_barrier ();
   3089   1.1  mrg 
   3090   1.1  mrg 
   3091   1.7  mrg File: libgomp.info,  Node: Implementing SINGLE construct,  Next: Implementing OpenACC's PARALLEL construct,  Prev: Implementing SECTIONS construct,  Up: The libgomp ABI
   3092   1.1  mrg 
   3093   1.7  mrg 9.14 Implementing SINGLE construct
   3094   1.1  mrg ==================================
   3095   1.1  mrg 
   3096   1.1  mrg A block like
   3097   1.1  mrg 
   3098   1.1  mrg        #pragma omp single
   3099   1.1  mrg        {
   3100   1.1  mrg          body;
   3101   1.1  mrg        }
   3102   1.1  mrg 
   3103   1.1  mrg    becomes
   3104   1.1  mrg 
   3105   1.1  mrg        if (GOMP_single_start ())
   3106   1.1  mrg          body;
   3107   1.1  mrg        GOMP_barrier ();
   3108   1.1  mrg 
   3109   1.1  mrg    while
   3110   1.1  mrg 
   3111   1.1  mrg        #pragma omp single copyprivate(x)
   3112   1.1  mrg          body;
   3113   1.1  mrg 
   3114   1.1  mrg    becomes
   3115   1.1  mrg 
   3116   1.1  mrg        datap = GOMP_single_copy_start ();
   3117   1.1  mrg        if (datap == NULL)
   3118   1.1  mrg          {
   3119   1.1  mrg            body;
   3120   1.1  mrg            data.x = x;
   3121   1.1  mrg            GOMP_single_copy_end (&data);
   3122   1.1  mrg          }
   3123   1.1  mrg        else
   3124   1.1  mrg          x = datap->x;
   3125   1.1  mrg        GOMP_barrier ();
   3126   1.1  mrg 
   3127   1.1  mrg 
   3128   1.7  mrg File: libgomp.info,  Node: Implementing OpenACC's PARALLEL construct,  Prev: Implementing SINGLE construct,  Up: The libgomp ABI
   3129   1.7  mrg 
   3130   1.7  mrg 9.15 Implementing OpenACC's PARALLEL construct
   3131   1.7  mrg ==============================================
   3132   1.7  mrg 
   3133   1.7  mrg        void GOACC_parallel ()
   3134   1.7  mrg 
   3135   1.7  mrg 
   3136   1.1  mrg File: libgomp.info,  Node: Reporting Bugs,  Next: Copying,  Prev: The libgomp ABI,  Up: Top
   3137   1.1  mrg 
   3138   1.7  mrg 10 Reporting Bugs
   3139   1.7  mrg *****************
   3140   1.1  mrg 
   3141   1.5  mrg Bugs in the GNU Offloading and Multi Processing Runtime Library should
   3142   1.5  mrg be reported via Bugzilla (http://gcc.gnu.org/bugzilla/).  Please add
   3143  1.11  mrg "openacc", or "openmp", or both to the keywords field in the bug report,
   3144  1.11  mrg as appropriate.
   3145   1.1  mrg 
   3146   1.1  mrg 
   3147   1.1  mrg File: libgomp.info,  Node: Copying,  Next: GNU Free Documentation License,  Prev: Reporting Bugs,  Up: Top
   3148   1.1  mrg 
   3149   1.3  mrg GNU General Public License
   3150   1.1  mrg **************************
   3151   1.1  mrg 
   3152   1.3  mrg                         Version 3, 29 June 2007
   3153   1.1  mrg 
   3154  1.11  mrg      Copyright (C) 2007 Free Software Foundation, Inc. <http://fsf.org/>
   3155   1.1  mrg 
   3156   1.3  mrg      Everyone is permitted to copy and distribute verbatim copies of this
   3157   1.3  mrg      license document, but changing it is not allowed.
   3158   1.1  mrg 
   3159   1.1  mrg Preamble
   3160   1.1  mrg ========
   3161   1.1  mrg 
   3162   1.3  mrg The GNU General Public License is a free, copyleft license for software
   3163   1.3  mrg and other kinds of works.
   3164   1.3  mrg 
   3165   1.3  mrg    The licenses for most software and other practical works are designed
   3166   1.3  mrg to take away your freedom to share and change the works.  By contrast,
   3167   1.3  mrg the GNU General Public License is intended to guarantee your freedom to
   3168   1.3  mrg share and change all versions of a program-to make sure it remains free
   3169   1.3  mrg software for all its users.  We, the Free Software Foundation, use the
   3170   1.3  mrg GNU General Public License for most of our software; it applies also to
   3171   1.3  mrg any other work released this way by its authors.  You can apply it to
   3172   1.1  mrg your programs, too.
   3173   1.1  mrg 
   3174   1.1  mrg    When we speak of free software, we are referring to freedom, not
   3175   1.1  mrg price.  Our General Public Licenses are designed to make sure that you
   3176   1.1  mrg have the freedom to distribute copies of free software (and charge for
   3177   1.3  mrg them if you wish), that you receive source code or can get it if you
   3178   1.3  mrg want it, that you can change the software or use pieces of it in new
   3179   1.3  mrg free programs, and that you know you can do these things.
   3180   1.3  mrg 
   3181   1.3  mrg    To protect your rights, we need to prevent others from denying you
   3182  1.11  mrg these rights or asking you to surrender the rights.  Therefore, you have
   3183  1.11  mrg certain responsibilities if you distribute copies of the software, or if
   3184  1.11  mrg you modify it: responsibilities to respect the freedom of others.
   3185   1.1  mrg 
   3186   1.1  mrg    For example, if you distribute copies of such a program, whether
   3187   1.3  mrg gratis or for a fee, you must pass on to the recipients the same
   3188   1.3  mrg freedoms that you received.  You must make sure that they, too, receive
   3189   1.3  mrg or can get the source code.  And you must show them these terms so they
   3190   1.3  mrg know their rights.
   3191   1.3  mrg 
   3192   1.3  mrg    Developers that use the GNU GPL protect your rights with two steps:
   3193   1.3  mrg (1) assert copyright on the software, and (2) offer you this License
   3194   1.3  mrg giving you legal permission to copy, distribute and/or modify it.
   3195   1.3  mrg 
   3196   1.3  mrg    For the developers' and authors' protection, the GPL clearly explains
   3197   1.3  mrg that there is no warranty for this free software.  For both users' and
   3198   1.3  mrg authors' sake, the GPL requires that modified versions be marked as
   3199   1.3  mrg changed, so that their problems will not be attributed erroneously to
   3200   1.3  mrg authors of previous versions.
   3201   1.3  mrg 
   3202   1.3  mrg    Some devices are designed to deny users access to install or run
   3203  1.11  mrg modified versions of the software inside them, although the manufacturer
   3204  1.11  mrg can do so.  This is fundamentally incompatible with the aim of
   3205  1.11  mrg protecting users' freedom to change the software.  The systematic
   3206  1.11  mrg pattern of such abuse occurs in the area of products for individuals to
   3207  1.11  mrg use, which is precisely where it is most unacceptable.  Therefore, we
   3208  1.11  mrg have designed this version of the GPL to prohibit the practice for those
   3209  1.11  mrg products.  If such problems arise substantially in other domains, we
   3210  1.11  mrg stand ready to extend this provision to those domains in future versions
   3211  1.11  mrg of the GPL, as needed to protect the freedom of users.
   3212   1.3  mrg 
   3213   1.3  mrg    Finally, every program is threatened constantly by software patents.
   3214   1.3  mrg States should not allow patents to restrict development and use of
   3215   1.3  mrg software on general-purpose computers, but in those that do, we wish to
   3216   1.3  mrg avoid the special danger that patents applied to a free program could
   3217   1.3  mrg make it effectively proprietary.  To prevent this, the GPL assures that
   3218   1.3  mrg patents cannot be used to render the program non-free.
   3219   1.1  mrg 
   3220   1.1  mrg    The precise terms and conditions for copying, distribution and
   3221   1.1  mrg modification follow.
   3222   1.1  mrg 
   3223   1.3  mrg TERMS AND CONDITIONS
   3224   1.3  mrg ====================
   3225   1.3  mrg 
   3226   1.3  mrg   0. Definitions.
   3227   1.3  mrg 
   3228   1.3  mrg      "This License" refers to version 3 of the GNU General Public
   3229   1.3  mrg      License.
   3230   1.1  mrg 
   3231   1.3  mrg      "Copyright" also means copyright-like laws that apply to other
   3232   1.3  mrg      kinds of works, such as semiconductor masks.
   3233   1.3  mrg 
   3234   1.3  mrg      "The Program" refers to any copyrightable work licensed under this
   3235   1.3  mrg      License.  Each licensee is addressed as "you".  "Licensees" and
   3236   1.3  mrg      "recipients" may be individuals or organizations.
   3237   1.3  mrg 
   3238   1.3  mrg      To "modify" a work means to copy from or adapt all or part of the
   3239   1.3  mrg      work in a fashion requiring copyright permission, other than the
   3240   1.3  mrg      making of an exact copy.  The resulting work is called a "modified
   3241   1.3  mrg      version" of the earlier work or a work "based on" the earlier work.
   3242   1.3  mrg 
   3243   1.3  mrg      A "covered work" means either the unmodified Program or a work
   3244   1.3  mrg      based on the Program.
   3245   1.3  mrg 
   3246   1.3  mrg      To "propagate" a work means to do anything with it that, without
   3247   1.3  mrg      permission, would make you directly or secondarily liable for
   3248  1.11  mrg      infringement under applicable copyright law, except executing it on
   3249  1.11  mrg      a computer or modifying a private copy.  Propagation includes
   3250   1.3  mrg      copying, distribution (with or without modification), making
   3251   1.3  mrg      available to the public, and in some countries other activities as
   3252   1.3  mrg      well.
   3253   1.3  mrg 
   3254   1.3  mrg      To "convey" a work means any kind of propagation that enables other
   3255   1.3  mrg      parties to make or receive copies.  Mere interaction with a user
   3256   1.3  mrg      through a computer network, with no transfer of a copy, is not
   3257   1.3  mrg      conveying.
   3258   1.3  mrg 
   3259   1.3  mrg      An interactive user interface displays "Appropriate Legal Notices"
   3260   1.3  mrg      to the extent that it includes a convenient and prominently visible
   3261   1.3  mrg      feature that (1) displays an appropriate copyright notice, and (2)
   3262   1.3  mrg      tells the user that there is no warranty for the work (except to
   3263  1.11  mrg      the extent that warranties are provided), that licensees may convey
   3264  1.11  mrg      the work under this License, and how to view a copy of this
   3265   1.3  mrg      License.  If the interface presents a list of user commands or
   3266   1.3  mrg      options, such as a menu, a prominent item in the list meets this
   3267   1.3  mrg      criterion.
   3268   1.3  mrg 
   3269   1.3  mrg   1. Source Code.
   3270   1.3  mrg 
   3271   1.3  mrg      The "source code" for a work means the preferred form of the work
   3272  1.11  mrg      for making modifications to it.  "Object code" means any non-source
   3273  1.11  mrg      form of a work.
   3274   1.3  mrg 
   3275   1.3  mrg      A "Standard Interface" means an interface that either is an
   3276   1.3  mrg      official standard defined by a recognized standards body, or, in
   3277   1.3  mrg      the case of interfaces specified for a particular programming
   3278   1.3  mrg      language, one that is widely used among developers working in that
   3279   1.3  mrg      language.
   3280   1.3  mrg 
   3281   1.3  mrg      The "System Libraries" of an executable work include anything,
   3282   1.3  mrg      other than the work as a whole, that (a) is included in the normal
   3283   1.3  mrg      form of packaging a Major Component, but which is not part of that
   3284  1.11  mrg      Major Component, and (b) serves only to enable use of the work with
   3285  1.11  mrg      that Major Component, or to implement a Standard Interface for
   3286  1.11  mrg      which an implementation is available to the public in source code
   3287  1.11  mrg      form.  A "Major Component", in this context, means a major
   3288   1.3  mrg      essential component (kernel, window system, and so on) of the
   3289   1.3  mrg      specific operating system (if any) on which the executable work
   3290   1.3  mrg      runs, or a compiler used to produce the work, or an object code
   3291   1.3  mrg      interpreter used to run it.
   3292   1.3  mrg 
   3293   1.3  mrg      The "Corresponding Source" for a work in object code form means all
   3294   1.3  mrg      the source code needed to generate, install, and (for an executable
   3295  1.11  mrg      work) run the object code and to modify the work, including scripts
   3296  1.11  mrg      to control those activities.  However, it does not include the
   3297  1.11  mrg      work's System Libraries, or general-purpose tools or generally
   3298   1.3  mrg      available free programs which are used unmodified in performing
   3299   1.3  mrg      those activities but which are not part of the work.  For example,
   3300  1.11  mrg      Corresponding Source includes interface definition files associated
   3301  1.11  mrg      with source files for the work, and the source code for shared
   3302  1.11  mrg      libraries and dynamically linked subprograms that the work is
   3303  1.11  mrg      specifically designed to require, such as by intimate data
   3304   1.3  mrg      communication or control flow between those subprograms and other
   3305   1.3  mrg      parts of the work.
   3306   1.3  mrg 
   3307   1.3  mrg      The Corresponding Source need not include anything that users can
   3308   1.3  mrg      regenerate automatically from other parts of the Corresponding
   3309   1.3  mrg      Source.
   3310   1.3  mrg 
   3311   1.3  mrg      The Corresponding Source for a work in source code form is that
   3312   1.3  mrg      same work.
   3313   1.3  mrg 
   3314   1.3  mrg   2. Basic Permissions.
   3315   1.3  mrg 
   3316   1.3  mrg      All rights granted under this License are granted for the term of
   3317   1.3  mrg      copyright on the Program, and are irrevocable provided the stated
   3318   1.3  mrg      conditions are met.  This License explicitly affirms your unlimited
   3319   1.3  mrg      permission to run the unmodified Program.  The output from running
   3320  1.11  mrg      a covered work is covered by this License only if the output, given
   3321  1.11  mrg      its content, constitutes a covered work.  This License acknowledges
   3322  1.11  mrg      your rights of fair use or other equivalent, as provided by
   3323  1.11  mrg      copyright law.
   3324   1.3  mrg 
   3325   1.3  mrg      You may make, run and propagate covered works that you do not
   3326   1.3  mrg      convey, without conditions so long as your license otherwise
   3327   1.3  mrg      remains in force.  You may convey covered works to others for the
   3328  1.11  mrg      sole purpose of having them make modifications exclusively for you,
   3329  1.11  mrg      or provide you with facilities for running those works, provided
   3330  1.11  mrg      that you comply with the terms of this License in conveying all
   3331  1.11  mrg      material for which you do not control copyright.  Those thus making
   3332  1.11  mrg      or running the covered works for you must do so exclusively on your
   3333  1.11  mrg      behalf, under your direction and control, on terms that prohibit
   3334  1.11  mrg      them from making any copies of your copyrighted material outside
   3335  1.11  mrg      their relationship with you.
   3336   1.3  mrg 
   3337   1.3  mrg      Conveying under any other circumstances is permitted solely under
   3338   1.3  mrg      the conditions stated below.  Sublicensing is not allowed; section
   3339   1.3  mrg      10 makes it unnecessary.
   3340   1.3  mrg 
   3341   1.3  mrg   3. Protecting Users' Legal Rights From Anti-Circumvention Law.
   3342   1.3  mrg 
   3343   1.3  mrg      No covered work shall be deemed part of an effective technological
   3344   1.3  mrg      measure under any applicable law fulfilling obligations under
   3345   1.3  mrg      article 11 of the WIPO copyright treaty adopted on 20 December
   3346   1.3  mrg      1996, or similar laws prohibiting or restricting circumvention of
   3347   1.3  mrg      such measures.
   3348   1.3  mrg 
   3349   1.3  mrg      When you convey a covered work, you waive any legal power to forbid
   3350   1.3  mrg      circumvention of technological measures to the extent such
   3351   1.3  mrg      circumvention is effected by exercising rights under this License
   3352  1.11  mrg      with respect to the covered work, and you disclaim any intention to
   3353  1.11  mrg      limit operation or modification of the work as a means of
   3354   1.3  mrg      enforcing, against the work's users, your or third parties' legal
   3355   1.3  mrg      rights to forbid circumvention of technological measures.
   3356   1.3  mrg 
   3357   1.3  mrg   4. Conveying Verbatim Copies.
   3358   1.3  mrg 
   3359   1.3  mrg      You may convey verbatim copies of the Program's source code as you
   3360   1.3  mrg      receive it, in any medium, provided that you conspicuously and
   3361   1.3  mrg      appropriately publish on each copy an appropriate copyright notice;
   3362   1.3  mrg      keep intact all notices stating that this License and any
   3363   1.3  mrg      non-permissive terms added in accord with section 7 apply to the
   3364   1.3  mrg      code; keep intact all notices of the absence of any warranty; and
   3365   1.3  mrg      give all recipients a copy of this License along with the Program.
   3366   1.3  mrg 
   3367   1.3  mrg      You may charge any price or no price for each copy that you convey,
   3368   1.3  mrg      and you may offer support or warranty protection for a fee.
   3369   1.3  mrg 
   3370   1.3  mrg   5. Conveying Modified Source Versions.
   3371   1.3  mrg 
   3372   1.3  mrg      You may convey a work based on the Program, or the modifications to
   3373   1.3  mrg      produce it from the Program, in the form of source code under the
   3374   1.3  mrg      terms of section 4, provided that you also meet all of these
   3375   1.3  mrg      conditions:
   3376   1.3  mrg 
   3377   1.3  mrg        a. The work must carry prominent notices stating that you
   3378   1.3  mrg           modified it, and giving a relevant date.
   3379   1.3  mrg 
   3380   1.3  mrg        b. The work must carry prominent notices stating that it is
   3381   1.3  mrg           released under this License and any conditions added under
   3382   1.3  mrg           section 7.  This requirement modifies the requirement in
   3383   1.3  mrg           section 4 to "keep intact all notices".
   3384   1.3  mrg 
   3385   1.3  mrg        c. You must license the entire work, as a whole, under this
   3386   1.3  mrg           License to anyone who comes into possession of a copy.  This
   3387   1.3  mrg           License will therefore apply, along with any applicable
   3388   1.3  mrg           section 7 additional terms, to the whole of the work, and all
   3389   1.3  mrg           its parts, regardless of how they are packaged.  This License
   3390   1.3  mrg           gives no permission to license the work in any other way, but
   3391   1.3  mrg           it does not invalidate such permission if you have separately
   3392   1.3  mrg           received it.
   3393   1.3  mrg 
   3394   1.3  mrg        d. If the work has interactive user interfaces, each must display
   3395   1.3  mrg           Appropriate Legal Notices; however, if the Program has
   3396   1.3  mrg           interactive interfaces that do not display Appropriate Legal
   3397   1.3  mrg           Notices, your work need not make them do so.
   3398   1.3  mrg 
   3399   1.3  mrg      A compilation of a covered work with other separate and independent
   3400   1.3  mrg      works, which are not by their nature extensions of the covered
   3401   1.3  mrg      work, and which are not combined with it such as to form a larger
   3402   1.3  mrg      program, in or on a volume of a storage or distribution medium, is
   3403   1.3  mrg      called an "aggregate" if the compilation and its resulting
   3404   1.3  mrg      copyright are not used to limit the access or legal rights of the
   3405   1.3  mrg      compilation's users beyond what the individual works permit.
   3406   1.3  mrg      Inclusion of a covered work in an aggregate does not cause this
   3407   1.3  mrg      License to apply to the other parts of the aggregate.
   3408   1.3  mrg 
   3409   1.3  mrg   6. Conveying Non-Source Forms.
   3410   1.3  mrg 
   3411   1.3  mrg      You may convey a covered work in object code form under the terms
   3412   1.3  mrg      of sections 4 and 5, provided that you also convey the
   3413   1.3  mrg      machine-readable Corresponding Source under the terms of this
   3414   1.3  mrg      License, in one of these ways:
   3415   1.3  mrg 
   3416   1.3  mrg        a. Convey the object code in, or embodied in, a physical product
   3417   1.3  mrg           (including a physical distribution medium), accompanied by the
   3418   1.3  mrg           Corresponding Source fixed on a durable physical medium
   3419   1.3  mrg           customarily used for software interchange.
   3420   1.3  mrg 
   3421   1.3  mrg        b. Convey the object code in, or embodied in, a physical product
   3422   1.3  mrg           (including a physical distribution medium), accompanied by a
   3423  1.11  mrg           written offer, valid for at least three years and valid for as
   3424  1.11  mrg           long as you offer spare parts or customer support for that
   3425   1.3  mrg           product model, to give anyone who possesses the object code
   3426   1.3  mrg           either (1) a copy of the Corresponding Source for all the
   3427   1.3  mrg           software in the product that is covered by this License, on a
   3428   1.3  mrg           durable physical medium customarily used for software
   3429   1.3  mrg           interchange, for a price no more than your reasonable cost of
   3430   1.3  mrg           physically performing this conveying of source, or (2) access
   3431   1.3  mrg           to copy the Corresponding Source from a network server at no
   3432   1.3  mrg           charge.
   3433   1.3  mrg 
   3434  1.11  mrg        c. Convey individual copies of the object code with a copy of the
   3435  1.11  mrg           written offer to provide the Corresponding Source.  This
   3436   1.3  mrg           alternative is allowed only occasionally and noncommercially,
   3437   1.3  mrg           and only if you received the object code with such an offer,
   3438   1.3  mrg           in accord with subsection 6b.
   3439   1.3  mrg 
   3440   1.3  mrg        d. Convey the object code by offering access from a designated
   3441  1.11  mrg           place (gratis or for a charge), and offer equivalent access to
   3442  1.11  mrg           the Corresponding Source in the same way through the same
   3443   1.3  mrg           place at no further charge.  You need not require recipients
   3444   1.3  mrg           to copy the Corresponding Source along with the object code.
   3445   1.3  mrg           If the place to copy the object code is a network server, the
   3446  1.11  mrg           Corresponding Source may be on a different server (operated by
   3447  1.11  mrg           you or a third party) that supports equivalent copying
   3448  1.11  mrg           facilities, provided you maintain clear directions next to the
   3449  1.11  mrg           object code saying where to find the Corresponding Source.
   3450   1.3  mrg           Regardless of what server hosts the Corresponding Source, you
   3451  1.11  mrg           remain obligated to ensure that it is available for as long as
   3452  1.11  mrg           needed to satisfy these requirements.
   3453   1.3  mrg 
   3454   1.3  mrg        e. Convey the object code using peer-to-peer transmission,
   3455   1.3  mrg           provided you inform other peers where the object code and
   3456   1.3  mrg           Corresponding Source of the work are being offered to the
   3457   1.3  mrg           general public at no charge under subsection 6d.
   3458   1.3  mrg 
   3459   1.3  mrg      A separable portion of the object code, whose source code is
   3460   1.3  mrg      excluded from the Corresponding Source as a System Library, need
   3461   1.3  mrg      not be included in conveying the object code work.
   3462   1.3  mrg 
   3463   1.3  mrg      A "User Product" is either (1) a "consumer product", which means
   3464   1.3  mrg      any tangible personal property which is normally used for personal,
   3465   1.3  mrg      family, or household purposes, or (2) anything designed or sold for
   3466  1.11  mrg      incorporation into a dwelling.  In determining whether a product is
   3467  1.11  mrg      a consumer product, doubtful cases shall be resolved in favor of
   3468   1.3  mrg      coverage.  For a particular product received by a particular user,
   3469   1.3  mrg      "normally used" refers to a typical or common use of that class of
   3470   1.3  mrg      product, regardless of the status of the particular user or of the
   3471   1.3  mrg      way in which the particular user actually uses, or expects or is
   3472   1.3  mrg      expected to use, the product.  A product is a consumer product
   3473   1.3  mrg      regardless of whether the product has substantial commercial,
   3474   1.3  mrg      industrial or non-consumer uses, unless such uses represent the
   3475   1.3  mrg      only significant mode of use of the product.
   3476   1.3  mrg 
   3477   1.3  mrg      "Installation Information" for a User Product means any methods,
   3478   1.3  mrg      procedures, authorization keys, or other information required to
   3479   1.3  mrg      install and execute modified versions of a covered work in that
   3480   1.3  mrg      User Product from a modified version of its Corresponding Source.
   3481   1.3  mrg      The information must suffice to ensure that the continued
   3482   1.3  mrg      functioning of the modified object code is in no case prevented or
   3483   1.3  mrg      interfered with solely because modification has been made.
   3484   1.3  mrg 
   3485   1.3  mrg      If you convey an object code work under this section in, or with,
   3486   1.3  mrg      or specifically for use in, a User Product, and the conveying
   3487   1.3  mrg      occurs as part of a transaction in which the right of possession
   3488   1.3  mrg      and use of the User Product is transferred to the recipient in
   3489   1.3  mrg      perpetuity or for a fixed term (regardless of how the transaction
   3490   1.3  mrg      is characterized), the Corresponding Source conveyed under this
   3491   1.3  mrg      section must be accompanied by the Installation Information.  But
   3492   1.3  mrg      this requirement does not apply if neither you nor any third party
   3493   1.3  mrg      retains the ability to install modified object code on the User
   3494   1.3  mrg      Product (for example, the work has been installed in ROM).
   3495   1.3  mrg 
   3496   1.3  mrg      The requirement to provide Installation Information does not
   3497   1.3  mrg      include a requirement to continue to provide support service,
   3498  1.11  mrg      warranty, or updates for a work that has been modified or installed
   3499  1.11  mrg      by the recipient, or for the User Product in which it has been
   3500  1.11  mrg      modified or installed.  Access to a network may be denied when the
   3501  1.11  mrg      modification itself materially and adversely affects the operation
   3502  1.11  mrg      of the network or violates the rules and protocols for
   3503   1.3  mrg      communication across the network.
   3504   1.3  mrg 
   3505   1.3  mrg      Corresponding Source conveyed, and Installation Information
   3506   1.3  mrg      provided, in accord with this section must be in a format that is
   3507   1.3  mrg      publicly documented (and with an implementation available to the
   3508   1.3  mrg      public in source code form), and must require no special password
   3509   1.3  mrg      or key for unpacking, reading or copying.
   3510   1.3  mrg 
   3511   1.3  mrg   7. Additional Terms.
   3512   1.3  mrg 
   3513   1.3  mrg      "Additional permissions" are terms that supplement the terms of
   3514   1.3  mrg      this License by making exceptions from one or more of its
   3515   1.3  mrg      conditions.  Additional permissions that are applicable to the
   3516   1.3  mrg      entire Program shall be treated as though they were included in
   3517   1.3  mrg      this License, to the extent that they are valid under applicable
   3518   1.3  mrg      law.  If additional permissions apply only to part of the Program,
   3519   1.3  mrg      that part may be used separately under those permissions, but the
   3520   1.3  mrg      entire Program remains governed by this License without regard to
   3521   1.3  mrg      the additional permissions.
   3522   1.3  mrg 
   3523   1.3  mrg      When you convey a copy of a covered work, you may at your option
   3524   1.3  mrg      remove any additional permissions from that copy, or from any part
   3525   1.3  mrg      of it.  (Additional permissions may be written to require their own
   3526   1.3  mrg      removal in certain cases when you modify the work.)  You may place
   3527   1.3  mrg      additional permissions on material, added by you to a covered work,
   3528   1.3  mrg      for which you have or can give appropriate copyright permission.
   3529   1.3  mrg 
   3530   1.3  mrg      Notwithstanding any other provision of this License, for material
   3531   1.3  mrg      you add to a covered work, you may (if authorized by the copyright
   3532  1.11  mrg      holders of that material) supplement the terms of this License with
   3533  1.11  mrg      terms:
   3534   1.3  mrg 
   3535   1.3  mrg        a. Disclaiming warranty or limiting liability differently from
   3536   1.3  mrg           the terms of sections 15 and 16 of this License; or
   3537   1.3  mrg 
   3538   1.3  mrg        b. Requiring preservation of specified reasonable legal notices
   3539   1.3  mrg           or author attributions in that material or in the Appropriate
   3540   1.3  mrg           Legal Notices displayed by works containing it; or
   3541   1.3  mrg 
   3542   1.3  mrg        c. Prohibiting misrepresentation of the origin of that material,
   3543  1.11  mrg           or requiring that modified versions of such material be marked
   3544  1.11  mrg           in reasonable ways as different from the original version; or
   3545   1.3  mrg 
   3546   1.3  mrg        d. Limiting the use for publicity purposes of names of licensors
   3547   1.3  mrg           or authors of the material; or
   3548   1.3  mrg 
   3549   1.3  mrg        e. Declining to grant rights under trademark law for use of some
   3550   1.3  mrg           trade names, trademarks, or service marks; or
   3551   1.3  mrg 
   3552   1.3  mrg        f. Requiring indemnification of licensors and authors of that
   3553   1.3  mrg           material by anyone who conveys the material (or modified
   3554   1.3  mrg           versions of it) with contractual assumptions of liability to
   3555   1.3  mrg           the recipient, for any liability that these contractual
   3556   1.3  mrg           assumptions directly impose on those licensors and authors.
   3557   1.3  mrg 
   3558   1.3  mrg      All other non-permissive additional terms are considered "further
   3559   1.3  mrg      restrictions" within the meaning of section 10.  If the Program as
   3560   1.3  mrg      you received it, or any part of it, contains a notice stating that
   3561   1.3  mrg      it is governed by this License along with a term that is a further
   3562   1.3  mrg      restriction, you may remove that term.  If a license document
   3563  1.11  mrg      contains a further restriction but permits relicensing or conveying
   3564  1.11  mrg      under this License, you may add to a covered work material governed
   3565  1.11  mrg      by the terms of that license document, provided that the further
   3566  1.11  mrg      restriction does not survive such relicensing or conveying.
   3567   1.3  mrg 
   3568   1.3  mrg      If you add terms to a covered work in accord with this section, you
   3569   1.3  mrg      must place, in the relevant source files, a statement of the
   3570   1.3  mrg      additional terms that apply to those files, or a notice indicating
   3571   1.3  mrg      where to find the applicable terms.
   3572   1.3  mrg 
   3573   1.3  mrg      Additional terms, permissive or non-permissive, may be stated in
   3574   1.3  mrg      the form of a separately written license, or stated as exceptions;
   3575   1.3  mrg      the above requirements apply either way.
   3576   1.3  mrg 
   3577   1.3  mrg   8. Termination.
   3578   1.3  mrg 
   3579   1.3  mrg      You may not propagate or modify a covered work except as expressly
   3580   1.3  mrg      provided under this License.  Any attempt otherwise to propagate or
   3581   1.3  mrg      modify it is void, and will automatically terminate your rights
   3582  1.11  mrg      under this License (including any patent licenses granted under the
   3583  1.11  mrg      third paragraph of section 11).
   3584   1.3  mrg 
   3585   1.3  mrg      However, if you cease all violation of this License, then your
   3586   1.3  mrg      license from a particular copyright holder is reinstated (a)
   3587  1.11  mrg      provisionally, unless and until the copyright holder explicitly and
   3588  1.11  mrg      finally terminates your license, and (b) permanently, if the
   3589   1.3  mrg      copyright holder fails to notify you of the violation by some
   3590   1.3  mrg      reasonable means prior to 60 days after the cessation.
   3591   1.3  mrg 
   3592   1.3  mrg      Moreover, your license from a particular copyright holder is
   3593   1.3  mrg      reinstated permanently if the copyright holder notifies you of the
   3594   1.3  mrg      violation by some reasonable means, this is the first time you have
   3595   1.3  mrg      received notice of violation of this License (for any work) from
   3596   1.3  mrg      that copyright holder, and you cure the violation prior to 30 days
   3597   1.3  mrg      after your receipt of the notice.
   3598   1.3  mrg 
   3599   1.3  mrg      Termination of your rights under this section does not terminate
   3600  1.11  mrg      the licenses of parties who have received copies or rights from you
   3601  1.11  mrg      under this License.  If your rights have been terminated and not
   3602  1.11  mrg      permanently reinstated, you do not qualify to receive new licenses
   3603  1.11  mrg      for the same material under section 10.
   3604   1.3  mrg 
   3605   1.3  mrg   9. Acceptance Not Required for Having Copies.
   3606   1.3  mrg 
   3607   1.3  mrg      You are not required to accept this License in order to receive or
   3608   1.3  mrg      run a copy of the Program.  Ancillary propagation of a covered work
   3609   1.3  mrg      occurring solely as a consequence of using peer-to-peer
   3610   1.3  mrg      transmission to receive a copy likewise does not require
   3611   1.3  mrg      acceptance.  However, nothing other than this License grants you
   3612   1.3  mrg      permission to propagate or modify any covered work.  These actions
   3613   1.3  mrg      infringe copyright if you do not accept this License.  Therefore,
   3614   1.3  mrg      by modifying or propagating a covered work, you indicate your
   3615   1.3  mrg      acceptance of this License to do so.
   3616   1.3  mrg 
   3617  1.11  mrg   10. Automatic Licensing of Downstream Recipients.
   3618   1.3  mrg 
   3619   1.3  mrg      Each time you convey a covered work, the recipient automatically
   3620   1.3  mrg      receives a license from the original licensors, to run, modify and
   3621   1.3  mrg      propagate that work, subject to this License.  You are not
   3622   1.3  mrg      responsible for enforcing compliance by third parties with this
   3623   1.3  mrg      License.
   3624   1.3  mrg 
   3625   1.3  mrg      An "entity transaction" is a transaction transferring control of an
   3626   1.3  mrg      organization, or substantially all assets of one, or subdividing an
   3627   1.3  mrg      organization, or merging organizations.  If propagation of a
   3628   1.3  mrg      covered work results from an entity transaction, each party to that
   3629   1.3  mrg      transaction who receives a copy of the work also receives whatever
   3630   1.3  mrg      licenses to the work the party's predecessor in interest had or
   3631  1.11  mrg      could give under the previous paragraph, plus a right to possession
   3632  1.11  mrg      of the Corresponding Source of the work from the predecessor in
   3633  1.11  mrg      interest, if the predecessor has it or can get it with reasonable
   3634  1.11  mrg      efforts.
   3635   1.3  mrg 
   3636   1.3  mrg      You may not impose any further restrictions on the exercise of the
   3637   1.3  mrg      rights granted or affirmed under this License.  For example, you
   3638  1.11  mrg      may not impose a license fee, royalty, or other charge for exercise
   3639  1.11  mrg      of rights granted under this License, and you may not initiate
   3640  1.11  mrg      litigation (including a cross-claim or counterclaim in a lawsuit)
   3641  1.11  mrg      alleging that any patent claim is infringed by making, using,
   3642  1.11  mrg      selling, offering for sale, or importing the Program or any portion
   3643  1.11  mrg      of it.
   3644   1.3  mrg 
   3645  1.11  mrg   11. Patents.
   3646   1.3  mrg 
   3647   1.3  mrg      A "contributor" is a copyright holder who authorizes use under this
   3648   1.3  mrg      License of the Program or a work on which the Program is based.
   3649   1.3  mrg      The work thus licensed is called the contributor's "contributor
   3650   1.3  mrg      version".
   3651   1.3  mrg 
   3652   1.3  mrg      A contributor's "essential patent claims" are all patent claims
   3653   1.3  mrg      owned or controlled by the contributor, whether already acquired or
   3654   1.3  mrg      hereafter acquired, that would be infringed by some manner,
   3655   1.3  mrg      permitted by this License, of making, using, or selling its
   3656   1.3  mrg      contributor version, but do not include claims that would be
   3657   1.3  mrg      infringed only as a consequence of further modification of the
   3658   1.3  mrg      contributor version.  For purposes of this definition, "control"
   3659   1.3  mrg      includes the right to grant patent sublicenses in a manner
   3660   1.3  mrg      consistent with the requirements of this License.
   3661   1.3  mrg 
   3662   1.3  mrg      Each contributor grants you a non-exclusive, worldwide,
   3663   1.3  mrg      royalty-free patent license under the contributor's essential
   3664   1.3  mrg      patent claims, to make, use, sell, offer for sale, import and
   3665  1.11  mrg      otherwise run, modify and propagate the contents of its contributor
   3666  1.11  mrg      version.
   3667   1.3  mrg 
   3668   1.3  mrg      In the following three paragraphs, a "patent license" is any
   3669   1.3  mrg      express agreement or commitment, however denominated, not to
   3670   1.3  mrg      enforce a patent (such as an express permission to practice a
   3671  1.11  mrg      patent or covenant not to sue for patent infringement).  To "grant"
   3672  1.11  mrg      such a patent license to a party means to make such an agreement or
   3673  1.11  mrg      commitment not to enforce a patent against the party.
   3674   1.3  mrg 
   3675   1.3  mrg      If you convey a covered work, knowingly relying on a patent
   3676   1.3  mrg      license, and the Corresponding Source of the work is not available
   3677   1.3  mrg      for anyone to copy, free of charge and under the terms of this
   3678   1.3  mrg      License, through a publicly available network server or other
   3679   1.3  mrg      readily accessible means, then you must either (1) cause the
   3680   1.3  mrg      Corresponding Source to be so available, or (2) arrange to deprive
   3681   1.3  mrg      yourself of the benefit of the patent license for this particular
   3682   1.3  mrg      work, or (3) arrange, in a manner consistent with the requirements
   3683   1.3  mrg      of this License, to extend the patent license to downstream
   3684   1.3  mrg      recipients.  "Knowingly relying" means you have actual knowledge
   3685   1.3  mrg      that, but for the patent license, your conveying the covered work
   3686   1.3  mrg      in a country, or your recipient's use of the covered work in a
   3687   1.3  mrg      country, would infringe one or more identifiable patents in that
   3688   1.3  mrg      country that you have reason to believe are valid.
   3689   1.3  mrg 
   3690   1.3  mrg      If, pursuant to or in connection with a single transaction or
   3691   1.3  mrg      arrangement, you convey, or propagate by procuring conveyance of, a
   3692   1.3  mrg      covered work, and grant a patent license to some of the parties
   3693   1.3  mrg      receiving the covered work authorizing them to use, propagate,
   3694   1.3  mrg      modify or convey a specific copy of the covered work, then the
   3695   1.3  mrg      patent license you grant is automatically extended to all
   3696   1.3  mrg      recipients of the covered work and works based on it.
   3697   1.3  mrg 
   3698   1.3  mrg      A patent license is "discriminatory" if it does not include within
   3699   1.3  mrg      the scope of its coverage, prohibits the exercise of, or is
   3700   1.3  mrg      conditioned on the non-exercise of one or more of the rights that
   3701   1.3  mrg      are specifically granted under this License.  You may not convey a
   3702   1.3  mrg      covered work if you are a party to an arrangement with a third
   3703  1.11  mrg      party that is in the business of distributing software, under which
   3704  1.11  mrg      you make payment to the third party based on the extent of your
   3705  1.11  mrg      activity of conveying the work, and under which the third party
   3706  1.11  mrg      grants, to any of the parties who would receive the covered work
   3707  1.11  mrg      from you, a discriminatory patent license (a) in connection with
   3708  1.11  mrg      copies of the covered work conveyed by you (or copies made from
   3709  1.11  mrg      those copies), or (b) primarily for and in connection with specific
   3710  1.11  mrg      products or compilations that contain the covered work, unless you
   3711  1.11  mrg      entered into that arrangement, or that patent license was granted,
   3712  1.11  mrg      prior to 28 March 2007.
   3713   1.3  mrg 
   3714   1.3  mrg      Nothing in this License shall be construed as excluding or limiting
   3715   1.3  mrg      any implied license or other defenses to infringement that may
   3716   1.3  mrg      otherwise be available to you under applicable patent law.
   3717   1.3  mrg 
   3718  1.11  mrg   12. No Surrender of Others' Freedom.
   3719   1.3  mrg 
   3720  1.11  mrg      If conditions are imposed on you (whether by court order, agreement
   3721  1.11  mrg      or otherwise) that contradict the conditions of this License, they
   3722  1.11  mrg      do not excuse you from the conditions of this License.  If you
   3723  1.11  mrg      cannot convey a covered work so as to satisfy simultaneously your
   3724  1.11  mrg      obligations under this License and any other pertinent obligations,
   3725  1.11  mrg      then as a consequence you may not convey it at all.  For example,
   3726  1.11  mrg      if you agree to terms that obligate you to collect a royalty for
   3727  1.11  mrg      further conveying from those to whom you convey the Program, the
   3728  1.11  mrg      only way you could satisfy both those terms and this License would
   3729  1.11  mrg      be to refrain entirely from conveying the Program.
   3730   1.3  mrg 
   3731  1.11  mrg   13. Use with the GNU Affero General Public License.
   3732   1.3  mrg 
   3733   1.3  mrg      Notwithstanding any other provision of this License, you have
   3734   1.3  mrg      permission to link or combine any covered work with a work licensed
   3735   1.3  mrg      under version 3 of the GNU Affero General Public License into a
   3736   1.3  mrg      single combined work, and to convey the resulting work.  The terms
   3737   1.3  mrg      of this License will continue to apply to the part which is the
   3738   1.3  mrg      covered work, but the special requirements of the GNU Affero
   3739   1.3  mrg      General Public License, section 13, concerning interaction through
   3740   1.3  mrg      a network will apply to the combination as such.
   3741   1.3  mrg 
   3742  1.11  mrg   14. Revised Versions of this License.
   3743   1.3  mrg 
   3744   1.3  mrg      The Free Software Foundation may publish revised and/or new
   3745  1.11  mrg      versions of the GNU General Public License from time to time.  Such
   3746  1.11  mrg      new versions will be similar in spirit to the present version, but
   3747  1.11  mrg      may differ in detail to address new problems or concerns.
   3748   1.1  mrg 
   3749   1.1  mrg      Each version is given a distinguishing version number.  If the
   3750   1.3  mrg      Program specifies that a certain numbered version of the GNU
   3751   1.3  mrg      General Public License "or any later version" applies to it, you
   3752   1.3  mrg      have the option of following the terms and conditions either of
   3753  1.11  mrg      that numbered version or of any later version published by the Free
   3754  1.11  mrg      Software Foundation.  If the Program does not specify a version
   3755  1.11  mrg      number of the GNU General Public License, you may choose any
   3756  1.11  mrg      version ever published by the Free Software Foundation.
   3757   1.1  mrg 
   3758   1.3  mrg      If the Program specifies that a proxy can decide which future
   3759   1.3  mrg      versions of the GNU General Public License can be used, that
   3760   1.3  mrg      proxy's public statement of acceptance of a version permanently
   3761   1.3  mrg      authorizes you to choose that version for the Program.
   3762   1.3  mrg 
   3763   1.3  mrg      Later license versions may give you additional or different
   3764   1.3  mrg      permissions.  However, no additional obligations are imposed on any
   3765   1.3  mrg      author or copyright holder as a result of your choosing to follow a
   3766   1.3  mrg      later version.
   3767   1.3  mrg 
   3768  1.11  mrg   15. Disclaimer of Warranty.
   3769   1.3  mrg 
   3770   1.3  mrg      THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
   3771  1.11  mrg      APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE
   3772   1.3  mrg      COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS"
   3773   1.3  mrg      WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED,
   3774   1.3  mrg      INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
   3775  1.11  mrg      MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE
   3776   1.3  mrg      RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.
   3777   1.3  mrg      SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL
   3778   1.3  mrg      NECESSARY SERVICING, REPAIR OR CORRECTION.
   3779   1.3  mrg 
   3780  1.11  mrg   16. Limitation of Liability.
   3781   1.3  mrg 
   3782   1.3  mrg      IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN
   3783   1.3  mrg      WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES
   3784  1.11  mrg      AND/OR CONVEYS THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR
   3785  1.11  mrg      DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR
   3786   1.3  mrg      CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE
   3787   1.3  mrg      THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA
   3788   1.3  mrg      BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
   3789   1.3  mrg      PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
   3790   1.3  mrg      PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF
   3791   1.3  mrg      THE POSSIBILITY OF SUCH DAMAGES.
   3792   1.3  mrg 
   3793  1.11  mrg   17. Interpretation of Sections 15 and 16.
   3794   1.3  mrg 
   3795   1.3  mrg      If the disclaimer of warranty and limitation of liability provided
   3796   1.3  mrg      above cannot be given local legal effect according to their terms,
   3797   1.3  mrg      reviewing courts shall apply local law that most closely
   3798   1.3  mrg      approximates an absolute waiver of all civil liability in
   3799   1.3  mrg      connection with the Program, unless a warranty or assumption of
   3800   1.3  mrg      liability accompanies a copy of the Program in return for a fee.
   3801   1.1  mrg 
   3802   1.3  mrg END OF TERMS AND CONDITIONS
   3803   1.3  mrg ===========================
   3804   1.3  mrg 
   3805   1.3  mrg How to Apply These Terms to Your New Programs
   3806   1.3  mrg =============================================
   3807   1.1  mrg 
   3808   1.1  mrg If you develop a new program, and you want it to be of the greatest
   3809   1.1  mrg possible use to the public, the best way to achieve this is to make it
   3810   1.1  mrg free software which everyone can redistribute and change under these
   3811   1.1  mrg terms.
   3812   1.1  mrg 
   3813   1.1  mrg    To do so, attach the following notices to the program.  It is safest
   3814   1.1  mrg to attach them to the start of each source file to most effectively
   3815   1.3  mrg state the exclusion of warranty; and each file should have at least the
   3816   1.3  mrg "copyright" line and a pointer to where the full notice is found.
   3817   1.1  mrg 
   3818   1.1  mrg      ONE LINE TO GIVE THE PROGRAM'S NAME AND A BRIEF IDEA OF WHAT IT DOES.
   3819   1.3  mrg      Copyright (C) YEAR NAME OF AUTHOR
   3820   1.1  mrg 
   3821   1.3  mrg      This program is free software: you can redistribute it and/or modify
   3822   1.1  mrg      it under the terms of the GNU General Public License as published by
   3823   1.3  mrg      the Free Software Foundation, either version 3 of the License, or (at
   3824   1.3  mrg      your option) any later version.
   3825   1.1  mrg 
   3826   1.3  mrg      This program is distributed in the hope that it will be useful, but
   3827   1.3  mrg      WITHOUT ANY WARRANTY; without even the implied warranty of
   3828   1.3  mrg      MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
   3829   1.3  mrg      General Public License for more details.
   3830   1.1  mrg 
   3831   1.1  mrg      You should have received a copy of the GNU General Public License
   3832  1.11  mrg      along with this program.  If not, see <http://www.gnu.org/licenses/>.
   3833   1.1  mrg 
   3834   1.1  mrg    Also add information on how to contact you by electronic and paper
   3835   1.1  mrg mail.
   3836   1.1  mrg 
   3837   1.3  mrg    If the program does terminal interaction, make it output a short
   3838   1.3  mrg notice like this when it starts in an interactive mode:
   3839   1.1  mrg 
   3840   1.3  mrg      PROGRAM Copyright (C) YEAR NAME OF AUTHOR
   3841  1.11  mrg      This program comes with ABSOLUTELY NO WARRANTY; for details type 'show w'.
   3842   1.1  mrg      This is free software, and you are welcome to redistribute it
   3843  1.11  mrg      under certain conditions; type 'show c' for details.
   3844   1.1  mrg 
   3845  1.11  mrg    The hypothetical commands 'show w' and 'show c' should show the
   3846   1.3  mrg appropriate parts of the General Public License.  Of course, your
   3847   1.3  mrg program's commands might be different; for a GUI interface, you would
   3848   1.3  mrg use an "about box".
   3849   1.1  mrg 
   3850   1.1  mrg    You should also get your employer (if you work as a programmer) or
   3851   1.3  mrg school, if any, to sign a "copyright disclaimer" for the program, if
   3852   1.3  mrg necessary.  For more information on this, and how to apply and follow
   3853  1.11  mrg the GNU GPL, see <http://www.gnu.org/licenses/>.
   3854   1.1  mrg 
   3855   1.3  mrg    The GNU General Public License does not permit incorporating your
   3856   1.1  mrg program into proprietary programs.  If your program is a subroutine
   3857   1.1  mrg library, you may consider it more useful to permit linking proprietary
   3858   1.1  mrg applications with the library.  If this is what you want to do, use the
   3859   1.3  mrg GNU Lesser General Public License instead of this License.  But first,
   3860  1.11  mrg please read <http://www.gnu.org/philosophy/why-not-lgpl.html>.
   3861   1.1  mrg 
   3862   1.1  mrg 
   3863   1.1  mrg File: libgomp.info,  Node: GNU Free Documentation License,  Next: Funding,  Prev: Copying,  Up: Top
   3864   1.1  mrg 
   3865   1.1  mrg GNU Free Documentation License
   3866   1.1  mrg ******************************
   3867   1.1  mrg 
   3868   1.3  mrg                      Version 1.3, 3 November 2008
   3869   1.1  mrg 
   3870   1.3  mrg      Copyright (C) 2000, 2001, 2002, 2007, 2008 Free Software Foundation, Inc.
   3871  1.11  mrg      <http://fsf.org/>
   3872   1.1  mrg 
   3873   1.1  mrg      Everyone is permitted to copy and distribute verbatim copies
   3874   1.1  mrg      of this license document, but changing it is not allowed.
   3875   1.1  mrg 
   3876   1.1  mrg   0. PREAMBLE
   3877   1.1  mrg 
   3878   1.1  mrg      The purpose of this License is to make a manual, textbook, or other
   3879   1.1  mrg      functional and useful document "free" in the sense of freedom: to
   3880   1.1  mrg      assure everyone the effective freedom to copy and redistribute it,
   3881   1.1  mrg      with or without modifying it, either commercially or
   3882   1.1  mrg      noncommercially.  Secondarily, this License preserves for the
   3883   1.1  mrg      author and publisher a way to get credit for their work, while not
   3884   1.1  mrg      being considered responsible for modifications made by others.
   3885   1.1  mrg 
   3886   1.1  mrg      This License is a kind of "copyleft", which means that derivative
   3887   1.1  mrg      works of the document must themselves be free in the same sense.
   3888   1.1  mrg      It complements the GNU General Public License, which is a copyleft
   3889   1.1  mrg      license designed for free software.
   3890   1.1  mrg 
   3891   1.1  mrg      We have designed this License in order to use it for manuals for
   3892   1.1  mrg      free software, because free software needs free documentation: a
   3893   1.1  mrg      free program should come with manuals providing the same freedoms
   3894   1.1  mrg      that the software does.  But this License is not limited to
   3895   1.1  mrg      software manuals; it can be used for any textual work, regardless
   3896  1.11  mrg      of subject matter or whether it is published as a printed book.  We
   3897  1.11  mrg      recommend this License principally for works whose purpose is
   3898   1.1  mrg      instruction or reference.
   3899   1.1  mrg 
   3900   1.1  mrg   1. APPLICABILITY AND DEFINITIONS
   3901   1.1  mrg 
   3902   1.1  mrg      This License applies to any manual or other work, in any medium,
   3903  1.11  mrg      that contains a notice placed by the copyright holder saying it can
   3904  1.11  mrg      be distributed under the terms of this License.  Such a notice
   3905   1.1  mrg      grants a world-wide, royalty-free license, unlimited in duration,
   3906   1.1  mrg      to use that work under the conditions stated herein.  The
   3907   1.1  mrg      "Document", below, refers to any such manual or work.  Any member
   3908  1.11  mrg      of the public is a licensee, and is addressed as "you".  You accept
   3909  1.11  mrg      the license if you copy, modify or distribute the work in a way
   3910  1.11  mrg      requiring permission under copyright law.
   3911   1.1  mrg 
   3912   1.1  mrg      A "Modified Version" of the Document means any work containing the
   3913   1.1  mrg      Document or a portion of it, either copied verbatim, or with
   3914   1.1  mrg      modifications and/or translated into another language.
   3915   1.1  mrg 
   3916   1.1  mrg      A "Secondary Section" is a named appendix or a front-matter section
   3917   1.1  mrg      of the Document that deals exclusively with the relationship of the
   3918   1.1  mrg      publishers or authors of the Document to the Document's overall
   3919   1.1  mrg      subject (or to related matters) and contains nothing that could
   3920   1.1  mrg      fall directly within that overall subject.  (Thus, if the Document
   3921   1.1  mrg      is in part a textbook of mathematics, a Secondary Section may not
   3922   1.1  mrg      explain any mathematics.)  The relationship could be a matter of
   3923   1.1  mrg      historical connection with the subject or with related matters, or
   3924   1.1  mrg      of legal, commercial, philosophical, ethical or political position
   3925   1.1  mrg      regarding them.
   3926   1.1  mrg 
   3927   1.1  mrg      The "Invariant Sections" are certain Secondary Sections whose
   3928  1.11  mrg      titles are designated, as being those of Invariant Sections, in the
   3929  1.11  mrg      notice that says that the Document is released under this License.
   3930  1.11  mrg      If a section does not fit the above definition of Secondary then it
   3931  1.11  mrg      is not allowed to be designated as Invariant.  The Document may
   3932  1.11  mrg      contain zero Invariant Sections.  If the Document does not identify
   3933  1.11  mrg      any Invariant Sections then there are none.
   3934   1.1  mrg 
   3935   1.1  mrg      The "Cover Texts" are certain short passages of text that are
   3936   1.1  mrg      listed, as Front-Cover Texts or Back-Cover Texts, in the notice
   3937   1.1  mrg      that says that the Document is released under this License.  A
   3938   1.1  mrg      Front-Cover Text may be at most 5 words, and a Back-Cover Text may
   3939   1.1  mrg      be at most 25 words.
   3940   1.1  mrg 
   3941   1.1  mrg      A "Transparent" copy of the Document means a machine-readable copy,
   3942   1.1  mrg      represented in a format whose specification is available to the
   3943   1.1  mrg      general public, that is suitable for revising the document
   3944  1.11  mrg      straightforwardly with generic text editors or (for images composed
   3945  1.11  mrg      of pixels) generic paint programs or (for drawings) some widely
   3946  1.11  mrg      available drawing editor, and that is suitable for input to text
   3947  1.11  mrg      formatters or for automatic translation to a variety of formats
   3948  1.11  mrg      suitable for input to text formatters.  A copy made in an otherwise
   3949  1.11  mrg      Transparent file format whose markup, or absence of markup, has
   3950  1.11  mrg      been arranged to thwart or discourage subsequent modification by
   3951  1.11  mrg      readers is not Transparent.  An image format is not Transparent if
   3952  1.11  mrg      used for any substantial amount of text.  A copy that is not
   3953  1.11  mrg      "Transparent" is called "Opaque".
   3954   1.1  mrg 
   3955   1.1  mrg      Examples of suitable formats for Transparent copies include plain
   3956   1.1  mrg      ASCII without markup, Texinfo input format, LaTeX input format,
   3957  1.11  mrg      SGML or XML using a publicly available DTD, and standard-conforming
   3958  1.11  mrg      simple HTML, PostScript or PDF designed for human modification.
   3959  1.11  mrg      Examples of transparent image formats include PNG, XCF and JPG.
   3960  1.11  mrg      Opaque formats include proprietary formats that can be read and
   3961  1.11  mrg      edited only by proprietary word processors, SGML or XML for which
   3962  1.11  mrg      the DTD and/or processing tools are not generally available, and
   3963  1.11  mrg      the machine-generated HTML, PostScript or PDF produced by some word
   3964  1.11  mrg      processors for output purposes only.
   3965   1.1  mrg 
   3966   1.1  mrg      The "Title Page" means, for a printed book, the title page itself,
   3967   1.1  mrg      plus such following pages as are needed to hold, legibly, the
   3968   1.1  mrg      material this License requires to appear in the title page.  For
   3969   1.1  mrg      works in formats which do not have any title page as such, "Title
   3970   1.1  mrg      Page" means the text near the most prominent appearance of the
   3971   1.1  mrg      work's title, preceding the beginning of the body of the text.
   3972   1.1  mrg 
   3973   1.3  mrg      The "publisher" means any person or entity that distributes copies
   3974   1.3  mrg      of the Document to the public.
   3975   1.3  mrg 
   3976   1.1  mrg      A section "Entitled XYZ" means a named subunit of the Document
   3977   1.1  mrg      whose title either is precisely XYZ or contains XYZ in parentheses
   3978   1.1  mrg      following text that translates XYZ in another language.  (Here XYZ
   3979   1.1  mrg      stands for a specific section name mentioned below, such as
   3980   1.1  mrg      "Acknowledgements", "Dedications", "Endorsements", or "History".)
   3981   1.1  mrg      To "Preserve the Title" of such a section when you modify the
   3982   1.1  mrg      Document means that it remains a section "Entitled XYZ" according
   3983   1.1  mrg      to this definition.
   3984   1.1  mrg 
   3985   1.1  mrg      The Document may include Warranty Disclaimers next to the notice
   3986   1.1  mrg      which states that this License applies to the Document.  These
   3987   1.1  mrg      Warranty Disclaimers are considered to be included by reference in
   3988   1.1  mrg      this License, but only as regards disclaiming warranties: any other
   3989   1.1  mrg      implication that these Warranty Disclaimers may have is void and
   3990   1.1  mrg      has no effect on the meaning of this License.
   3991   1.1  mrg 
   3992   1.1  mrg   2. VERBATIM COPYING
   3993   1.1  mrg 
   3994   1.1  mrg      You may copy and distribute the Document in any medium, either
   3995   1.1  mrg      commercially or noncommercially, provided that this License, the
   3996   1.1  mrg      copyright notices, and the license notice saying this License
   3997   1.1  mrg      applies to the Document are reproduced in all copies, and that you
   3998   1.1  mrg      add no other conditions whatsoever to those of this License.  You
   3999   1.1  mrg      may not use technical measures to obstruct or control the reading
   4000   1.1  mrg      or further copying of the copies you make or distribute.  However,
   4001   1.1  mrg      you may accept compensation in exchange for copies.  If you
   4002  1.11  mrg      distribute a large enough number of copies you must also follow the
   4003  1.11  mrg      conditions in section 3.
   4004   1.1  mrg 
   4005   1.1  mrg      You may also lend copies, under the same conditions stated above,
   4006   1.1  mrg      and you may publicly display copies.
   4007   1.1  mrg 
   4008   1.1  mrg   3. COPYING IN QUANTITY
   4009   1.1  mrg 
   4010   1.1  mrg      If you publish printed copies (or copies in media that commonly
   4011   1.1  mrg      have printed covers) of the Document, numbering more than 100, and
   4012   1.1  mrg      the Document's license notice requires Cover Texts, you must
   4013   1.1  mrg      enclose the copies in covers that carry, clearly and legibly, all
   4014   1.1  mrg      these Cover Texts: Front-Cover Texts on the front cover, and
   4015   1.1  mrg      Back-Cover Texts on the back cover.  Both covers must also clearly
   4016   1.1  mrg      and legibly identify you as the publisher of these copies.  The
   4017  1.11  mrg      front cover must present the full title with all words of the title
   4018  1.11  mrg      equally prominent and visible.  You may add other material on the
   4019  1.11  mrg      covers in addition.  Copying with changes limited to the covers, as
   4020  1.11  mrg      long as they preserve the title of the Document and satisfy these
   4021  1.11  mrg      conditions, can be treated as verbatim copying in other respects.
   4022   1.1  mrg 
   4023   1.1  mrg      If the required texts for either cover are too voluminous to fit
   4024   1.1  mrg      legibly, you should put the first ones listed (as many as fit
   4025   1.1  mrg      reasonably) on the actual cover, and continue the rest onto
   4026   1.1  mrg      adjacent pages.
   4027   1.1  mrg 
   4028   1.1  mrg      If you publish or distribute Opaque copies of the Document
   4029  1.11  mrg      numbering more than 100, you must either include a machine-readable
   4030  1.11  mrg      Transparent copy along with each Opaque copy, or state in or with
   4031  1.11  mrg      each Opaque copy a computer-network location from which the general
   4032  1.11  mrg      network-using public has access to download using public-standard
   4033  1.11  mrg      network protocols a complete Transparent copy of the Document, free
   4034  1.11  mrg      of added material.  If you use the latter option, you must take
   4035  1.11  mrg      reasonably prudent steps, when you begin distribution of Opaque
   4036  1.11  mrg      copies in quantity, to ensure that this Transparent copy will
   4037  1.11  mrg      remain thus accessible at the stated location until at least one
   4038  1.11  mrg      year after the last time you distribute an Opaque copy (directly or
   4039  1.11  mrg      through your agents or retailers) of that edition to the public.
   4040   1.1  mrg 
   4041   1.1  mrg      It is requested, but not required, that you contact the authors of
   4042  1.11  mrg      the Document well before redistributing any large number of copies,
   4043  1.11  mrg      to give them a chance to provide you with an updated version of the
   4044  1.11  mrg      Document.
   4045   1.1  mrg 
   4046   1.1  mrg   4. MODIFICATIONS
   4047   1.1  mrg 
   4048   1.1  mrg      You may copy and distribute a Modified Version of the Document
   4049   1.1  mrg      under the conditions of sections 2 and 3 above, provided that you
   4050  1.11  mrg      release the Modified Version under precisely this License, with the
   4051  1.11  mrg      Modified Version filling the role of the Document, thus licensing
   4052  1.11  mrg      distribution and modification of the Modified Version to whoever
   4053  1.11  mrg      possesses a copy of it.  In addition, you must do these things in
   4054  1.11  mrg      the Modified Version:
   4055   1.1  mrg 
   4056   1.1  mrg        A. Use in the Title Page (and on the covers, if any) a title
   4057  1.11  mrg           distinct from that of the Document, and from those of previous
   4058  1.11  mrg           versions (which should, if there were any, be listed in the
   4059  1.11  mrg           History section of the Document).  You may use the same title
   4060  1.11  mrg           as a previous version if the original publisher of that
   4061  1.11  mrg           version gives permission.
   4062   1.1  mrg 
   4063   1.1  mrg        B. List on the Title Page, as authors, one or more persons or
   4064   1.1  mrg           entities responsible for authorship of the modifications in
   4065   1.1  mrg           the Modified Version, together with at least five of the
   4066   1.1  mrg           principal authors of the Document (all of its principal
   4067   1.1  mrg           authors, if it has fewer than five), unless they release you
   4068   1.1  mrg           from this requirement.
   4069   1.1  mrg 
   4070   1.1  mrg        C. State on the Title page the name of the publisher of the
   4071   1.1  mrg           Modified Version, as the publisher.
   4072   1.1  mrg 
   4073   1.1  mrg        D. Preserve all the copyright notices of the Document.
   4074   1.1  mrg 
   4075   1.1  mrg        E. Add an appropriate copyright notice for your modifications
   4076   1.1  mrg           adjacent to the other copyright notices.
   4077   1.1  mrg 
   4078   1.1  mrg        F. Include, immediately after the copyright notices, a license
   4079   1.1  mrg           notice giving the public permission to use the Modified
   4080   1.1  mrg           Version under the terms of this License, in the form shown in
   4081   1.1  mrg           the Addendum below.
   4082   1.1  mrg 
   4083   1.1  mrg        G. Preserve in that license notice the full lists of Invariant
   4084   1.1  mrg           Sections and required Cover Texts given in the Document's
   4085   1.1  mrg           license notice.
   4086   1.1  mrg 
   4087   1.1  mrg        H. Include an unaltered copy of this License.
   4088   1.1  mrg 
   4089   1.1  mrg        I. Preserve the section Entitled "History", Preserve its Title,
   4090   1.1  mrg           and add to it an item stating at least the title, year, new
   4091  1.11  mrg           authors, and publisher of the Modified Version as given on the
   4092  1.11  mrg           Title Page.  If there is no section Entitled "History" in the
   4093  1.11  mrg           Document, create one stating the title, year, authors, and
   4094  1.11  mrg           publisher of the Document as given on its Title Page, then add
   4095  1.11  mrg           an item describing the Modified Version as stated in the
   4096  1.11  mrg           previous sentence.
   4097   1.1  mrg 
   4098   1.1  mrg        J. Preserve the network location, if any, given in the Document
   4099   1.1  mrg           for public access to a Transparent copy of the Document, and
   4100   1.1  mrg           likewise the network locations given in the Document for
   4101  1.11  mrg           previous versions it was based on.  These may be placed in the
   4102  1.11  mrg           "History" section.  You may omit a network location for a work
   4103  1.11  mrg           that was published at least four years before the Document
   4104  1.11  mrg           itself, or if the original publisher of the version it refers
   4105  1.11  mrg           to gives permission.
   4106   1.1  mrg 
   4107   1.1  mrg        K. For any section Entitled "Acknowledgements" or "Dedications",
   4108  1.11  mrg           Preserve the Title of the section, and preserve in the section
   4109  1.11  mrg           all the substance and tone of each of the contributor
   4110   1.1  mrg           acknowledgements and/or dedications given therein.
   4111   1.1  mrg 
   4112  1.11  mrg        L. Preserve all the Invariant Sections of the Document, unaltered
   4113  1.11  mrg           in their text and in their titles.  Section numbers or the
   4114  1.11  mrg           equivalent are not considered part of the section titles.
   4115   1.1  mrg 
   4116   1.1  mrg        M. Delete any section Entitled "Endorsements".  Such a section
   4117   1.1  mrg           may not be included in the Modified Version.
   4118   1.1  mrg 
   4119   1.1  mrg        N. Do not retitle any existing section to be Entitled
   4120   1.1  mrg           "Endorsements" or to conflict in title with any Invariant
   4121   1.1  mrg           Section.
   4122   1.1  mrg 
   4123   1.1  mrg        O. Preserve any Warranty Disclaimers.
   4124   1.1  mrg 
   4125   1.1  mrg      If the Modified Version includes new front-matter sections or
   4126   1.1  mrg      appendices that qualify as Secondary Sections and contain no
   4127  1.11  mrg      material copied from the Document, you may at your option designate
   4128  1.11  mrg      some or all of these sections as invariant.  To do this, add their
   4129  1.11  mrg      titles to the list of Invariant Sections in the Modified Version's
   4130  1.11  mrg      license notice.  These titles must be distinct from any other
   4131  1.11  mrg      section titles.
   4132   1.1  mrg 
   4133   1.1  mrg      You may add a section Entitled "Endorsements", provided it contains
   4134   1.1  mrg      nothing but endorsements of your Modified Version by various
   4135   1.1  mrg      parties--for example, statements of peer review or that the text
   4136   1.1  mrg      has been approved by an organization as the authoritative
   4137   1.1  mrg      definition of a standard.
   4138   1.1  mrg 
   4139   1.1  mrg      You may add a passage of up to five words as a Front-Cover Text,
   4140  1.11  mrg      and a passage of up to 25 words as a Back-Cover Text, to the end of
   4141  1.11  mrg      the list of Cover Texts in the Modified Version.  Only one passage
   4142  1.11  mrg      of Front-Cover Text and one of Back-Cover Text may be added by (or
   4143  1.11  mrg      through arrangements made by) any one entity.  If the Document
   4144  1.11  mrg      already includes a cover text for the same cover, previously added
   4145  1.11  mrg      by you or by arrangement made by the same entity you are acting on
   4146  1.11  mrg      behalf of, you may not add another; but you may replace the old
   4147  1.11  mrg      one, on explicit permission from the previous publisher that added
   4148  1.11  mrg      the old one.
   4149   1.1  mrg 
   4150   1.1  mrg      The author(s) and publisher(s) of the Document do not by this
   4151   1.1  mrg      License give permission to use their names for publicity for or to
   4152   1.1  mrg      assert or imply endorsement of any Modified Version.
   4153   1.1  mrg 
   4154   1.1  mrg   5. COMBINING DOCUMENTS
   4155   1.1  mrg 
   4156   1.1  mrg      You may combine the Document with other documents released under
   4157   1.1  mrg      this License, under the terms defined in section 4 above for
   4158  1.11  mrg      modified versions, provided that you include in the combination all
   4159  1.11  mrg      of the Invariant Sections of all of the original documents,
   4160   1.1  mrg      unmodified, and list them all as Invariant Sections of your
   4161   1.1  mrg      combined work in its license notice, and that you preserve all
   4162   1.1  mrg      their Warranty Disclaimers.
   4163   1.1  mrg 
   4164   1.1  mrg      The combined work need only contain one copy of this License, and
   4165   1.1  mrg      multiple identical Invariant Sections may be replaced with a single
   4166   1.1  mrg      copy.  If there are multiple Invariant Sections with the same name
   4167   1.1  mrg      but different contents, make the title of each such section unique
   4168   1.1  mrg      by adding at the end of it, in parentheses, the name of the
   4169   1.1  mrg      original author or publisher of that section if known, or else a
   4170   1.1  mrg      unique number.  Make the same adjustment to the section titles in
   4171   1.1  mrg      the list of Invariant Sections in the license notice of the
   4172   1.1  mrg      combined work.
   4173   1.1  mrg 
   4174   1.1  mrg      In the combination, you must combine any sections Entitled
   4175   1.1  mrg      "History" in the various original documents, forming one section
   4176   1.1  mrg      Entitled "History"; likewise combine any sections Entitled
   4177   1.1  mrg      "Acknowledgements", and any sections Entitled "Dedications".  You
   4178   1.1  mrg      must delete all sections Entitled "Endorsements."
   4179   1.1  mrg 
   4180   1.1  mrg   6. COLLECTIONS OF DOCUMENTS
   4181   1.1  mrg 
   4182   1.1  mrg      You may make a collection consisting of the Document and other
   4183   1.1  mrg      documents released under this License, and replace the individual
   4184   1.1  mrg      copies of this License in the various documents with a single copy
   4185   1.1  mrg      that is included in the collection, provided that you follow the
   4186  1.11  mrg      rules of this License for verbatim copying of each of the documents
   4187  1.11  mrg      in all other respects.
   4188   1.1  mrg 
   4189   1.1  mrg      You may extract a single document from such a collection, and
   4190   1.1  mrg      distribute it individually under this License, provided you insert
   4191  1.11  mrg      a copy of this License into the extracted document, and follow this
   4192  1.11  mrg      License in all other respects regarding verbatim copying of that
   4193  1.11  mrg      document.
   4194   1.1  mrg 
   4195   1.1  mrg   7. AGGREGATION WITH INDEPENDENT WORKS
   4196   1.1  mrg 
   4197   1.1  mrg      A compilation of the Document or its derivatives with other
   4198  1.11  mrg      separate and independent documents or works, in or on a volume of a
   4199  1.11  mrg      storage or distribution medium, is called an "aggregate" if the
   4200   1.1  mrg      copyright resulting from the compilation is not used to limit the
   4201   1.1  mrg      legal rights of the compilation's users beyond what the individual
   4202   1.1  mrg      works permit.  When the Document is included in an aggregate, this
   4203   1.1  mrg      License does not apply to the other works in the aggregate which
   4204   1.1  mrg      are not themselves derivative works of the Document.
   4205   1.1  mrg 
   4206   1.1  mrg      If the Cover Text requirement of section 3 is applicable to these
   4207   1.1  mrg      copies of the Document, then if the Document is less than one half
   4208   1.1  mrg      of the entire aggregate, the Document's Cover Texts may be placed
   4209   1.1  mrg      on covers that bracket the Document within the aggregate, or the
   4210   1.1  mrg      electronic equivalent of covers if the Document is in electronic
   4211   1.1  mrg      form.  Otherwise they must appear on printed covers that bracket
   4212   1.1  mrg      the whole aggregate.
   4213   1.1  mrg 
   4214   1.1  mrg   8. TRANSLATION
   4215   1.1  mrg 
   4216   1.1  mrg      Translation is considered a kind of modification, so you may
   4217   1.1  mrg      distribute translations of the Document under the terms of section
   4218   1.1  mrg      4.  Replacing Invariant Sections with translations requires special
   4219   1.1  mrg      permission from their copyright holders, but you may include
   4220   1.1  mrg      translations of some or all Invariant Sections in addition to the
   4221   1.1  mrg      original versions of these Invariant Sections.  You may include a
   4222   1.1  mrg      translation of this License, and all the license notices in the
   4223   1.1  mrg      Document, and any Warranty Disclaimers, provided that you also
   4224   1.1  mrg      include the original English version of this License and the
   4225   1.1  mrg      original versions of those notices and disclaimers.  In case of a
   4226   1.1  mrg      disagreement between the translation and the original version of
   4227   1.1  mrg      this License or a notice or disclaimer, the original version will
   4228   1.1  mrg      prevail.
   4229   1.1  mrg 
   4230   1.1  mrg      If a section in the Document is Entitled "Acknowledgements",
   4231   1.1  mrg      "Dedications", or "History", the requirement (section 4) to
   4232   1.1  mrg      Preserve its Title (section 1) will typically require changing the
   4233   1.1  mrg      actual title.
   4234   1.1  mrg 
   4235   1.1  mrg   9. TERMINATION
   4236   1.1  mrg 
   4237   1.1  mrg      You may not copy, modify, sublicense, or distribute the Document
   4238   1.3  mrg      except as expressly provided under this License.  Any attempt
   4239   1.3  mrg      otherwise to copy, modify, sublicense, or distribute it is void,
   4240   1.3  mrg      and will automatically terminate your rights under this License.
   4241   1.3  mrg 
   4242   1.3  mrg      However, if you cease all violation of this License, then your
   4243   1.3  mrg      license from a particular copyright holder is reinstated (a)
   4244  1.11  mrg      provisionally, unless and until the copyright holder explicitly and
   4245  1.11  mrg      finally terminates your license, and (b) permanently, if the
   4246   1.3  mrg      copyright holder fails to notify you of the violation by some
   4247   1.3  mrg      reasonable means prior to 60 days after the cessation.
   4248   1.3  mrg 
   4249   1.3  mrg      Moreover, your license from a particular copyright holder is
   4250   1.3  mrg      reinstated permanently if the copyright holder notifies you of the
   4251   1.3  mrg      violation by some reasonable means, this is the first time you have
   4252   1.3  mrg      received notice of violation of this License (for any work) from
   4253   1.3  mrg      that copyright holder, and you cure the violation prior to 30 days
   4254   1.3  mrg      after your receipt of the notice.
   4255   1.3  mrg 
   4256   1.3  mrg      Termination of your rights under this section does not terminate
   4257  1.11  mrg      the licenses of parties who have received copies or rights from you
   4258  1.11  mrg      under this License.  If your rights have been terminated and not
   4259  1.11  mrg      permanently reinstated, receipt of a copy of some or all of the
   4260  1.11  mrg      same material does not give you any rights to use it.
   4261   1.1  mrg 
   4262  1.11  mrg   10. FUTURE REVISIONS OF THIS LICENSE
   4263   1.1  mrg 
   4264   1.1  mrg      The Free Software Foundation may publish new, revised versions of
   4265   1.1  mrg      the GNU Free Documentation License from time to time.  Such new
   4266   1.1  mrg      versions will be similar in spirit to the present version, but may
   4267   1.1  mrg      differ in detail to address new problems or concerns.  See
   4268  1.11  mrg      <http://www.gnu.org/copyleft/>.
   4269   1.1  mrg 
   4270   1.1  mrg      Each version of the License is given a distinguishing version
   4271   1.1  mrg      number.  If the Document specifies that a particular numbered
   4272   1.1  mrg      version of this License "or any later version" applies to it, you
   4273   1.1  mrg      have the option of following the terms and conditions either of
   4274   1.1  mrg      that specified version or of any later version that has been
   4275  1.11  mrg      published (not as a draft) by the Free Software Foundation.  If the
   4276  1.11  mrg      Document does not specify a version number of this License, you may
   4277  1.11  mrg      choose any version ever published (not as a draft) by the Free
   4278  1.11  mrg      Software Foundation.  If the Document specifies that a proxy can
   4279  1.11  mrg      decide which future versions of this License can be used, that
   4280   1.3  mrg      proxy's public statement of acceptance of a version permanently
   4281   1.3  mrg      authorizes you to choose that version for the Document.
   4282   1.3  mrg 
   4283  1.11  mrg   11. RELICENSING
   4284   1.3  mrg 
   4285   1.3  mrg      "Massive Multiauthor Collaboration Site" (or "MMC Site") means any
   4286   1.3  mrg      World Wide Web server that publishes copyrightable works and also
   4287   1.3  mrg      provides prominent facilities for anybody to edit those works.  A
   4288   1.3  mrg      public wiki that anybody can edit is an example of such a server.
   4289   1.3  mrg      A "Massive Multiauthor Collaboration" (or "MMC") contained in the
   4290   1.3  mrg      site means any set of copyrightable works thus published on the MMC
   4291   1.3  mrg      site.
   4292   1.3  mrg 
   4293   1.3  mrg      "CC-BY-SA" means the Creative Commons Attribution-Share Alike 3.0
   4294   1.3  mrg      license published by Creative Commons Corporation, a not-for-profit
   4295   1.3  mrg      corporation with a principal place of business in San Francisco,
   4296   1.3  mrg      California, as well as future copyleft versions of that license
   4297   1.3  mrg      published by that same organization.
   4298   1.3  mrg 
   4299   1.3  mrg      "Incorporate" means to publish or republish a Document, in whole or
   4300   1.3  mrg      in part, as part of another Document.
   4301   1.3  mrg 
   4302   1.3  mrg      An MMC is "eligible for relicensing" if it is licensed under this
   4303   1.3  mrg      License, and if all works that were first published under this
   4304   1.3  mrg      License somewhere other than this MMC, and subsequently
   4305   1.3  mrg      incorporated in whole or in part into the MMC, (1) had no cover
   4306   1.3  mrg      texts or invariant sections, and (2) were thus incorporated prior
   4307   1.3  mrg      to November 1, 2008.
   4308   1.3  mrg 
   4309   1.3  mrg      The operator of an MMC Site may republish an MMC contained in the
   4310   1.3  mrg      site under CC-BY-SA on the same site at any time before August 1,
   4311   1.3  mrg      2009, provided the MMC is eligible for relicensing.
   4312   1.3  mrg 
   4313   1.1  mrg ADDENDUM: How to use this License for your documents
   4314   1.1  mrg ====================================================
   4315   1.1  mrg 
   4316   1.1  mrg To use this License in a document you have written, include a copy of
   4317   1.1  mrg the License in the document and put the following copyright and license
   4318   1.1  mrg notices just after the title page:
   4319   1.1  mrg 
   4320   1.1  mrg        Copyright (C)  YEAR  YOUR NAME.
   4321   1.1  mrg        Permission is granted to copy, distribute and/or modify this document
   4322   1.3  mrg        under the terms of the GNU Free Documentation License, Version 1.3
   4323   1.1  mrg        or any later version published by the Free Software Foundation;
   4324   1.1  mrg        with no Invariant Sections, no Front-Cover Texts, and no Back-Cover
   4325   1.1  mrg        Texts.  A copy of the license is included in the section entitled ``GNU
   4326   1.1  mrg        Free Documentation License''.
   4327   1.1  mrg 
   4328   1.1  mrg    If you have Invariant Sections, Front-Cover Texts and Back-Cover
   4329  1.11  mrg Texts, replace the "with...Texts."  line with this:
   4330   1.1  mrg 
   4331   1.1  mrg          with the Invariant Sections being LIST THEIR TITLES, with
   4332   1.1  mrg          the Front-Cover Texts being LIST, and with the Back-Cover Texts
   4333   1.1  mrg          being LIST.
   4334   1.1  mrg 
   4335   1.1  mrg    If you have Invariant Sections without Cover Texts, or some other
   4336   1.1  mrg combination of the three, merge those two alternatives to suit the
   4337   1.1  mrg situation.
   4338   1.1  mrg 
   4339   1.1  mrg    If your document contains nontrivial examples of program code, we
   4340  1.11  mrg recommend releasing these examples in parallel under your choice of free
   4341  1.11  mrg software license, such as the GNU General Public License, to permit
   4342  1.11  mrg their use in free software.
   4343   1.1  mrg 
   4344   1.1  mrg 
   4345   1.3  mrg File: libgomp.info,  Node: Funding,  Next: Library Index,  Prev: GNU Free Documentation License,  Up: Top
   4346   1.1  mrg 
   4347   1.1  mrg Funding Free Software
   4348   1.1  mrg *********************
   4349   1.1  mrg 
   4350   1.1  mrg If you want to have more free software a few years from now, it makes
   4351   1.1  mrg sense for you to help encourage people to contribute funds for its
   4352   1.1  mrg development.  The most effective approach known is to encourage
   4353   1.1  mrg commercial redistributors to donate.
   4354   1.1  mrg 
   4355   1.1  mrg    Users of free software systems can boost the pace of development by
   4356   1.1  mrg encouraging for-a-fee distributors to donate part of their selling price
   4357   1.1  mrg to free software developers--the Free Software Foundation, and others.
   4358   1.1  mrg 
   4359   1.1  mrg    The way to convince distributors to do this is to demand it and
   4360   1.1  mrg expect it from them.  So when you compare distributors, judge them
   4361   1.1  mrg partly by how much they give to free software development.  Show
   4362   1.1  mrg distributors they must compete to be the one who gives the most.
   4363   1.1  mrg 
   4364   1.1  mrg    To make this approach work, you must insist on numbers that you can
   4365   1.1  mrg compare, such as, "We will donate ten dollars to the Frobnitz project
   4366   1.1  mrg for each disk sold."  Don't be satisfied with a vague promise, such as
   4367   1.1  mrg "A portion of the profits are donated," since it doesn't give a basis
   4368   1.1  mrg for comparison.
   4369   1.1  mrg 
   4370   1.1  mrg    Even a precise fraction "of the profits from this disk" is not very
   4371   1.1  mrg meaningful, since creative accounting and unrelated business decisions
   4372  1.11  mrg can greatly alter what fraction of the sales price counts as profit.  If
   4373  1.11  mrg the price you pay is $50, ten percent of the profit is probably less
   4374   1.1  mrg than a dollar; it might be a few cents, or nothing at all.
   4375   1.1  mrg 
   4376   1.1  mrg    Some redistributors do development work themselves.  This is useful
   4377   1.1  mrg too; but to keep everyone honest, you need to inquire how much they do,
   4378   1.1  mrg and what kind.  Some kinds of development make much more long-term
   4379   1.1  mrg difference than others.  For example, maintaining a separate version of
   4380   1.1  mrg a program contributes very little; maintaining the standard version of a
   4381   1.1  mrg program for the whole community contributes much.  Easy new ports
   4382   1.1  mrg contribute little, since someone else would surely do them; difficult
   4383  1.11  mrg ports such as adding a new CPU to the GNU Compiler Collection contribute
   4384  1.11  mrg more; major new features or packages contribute the most.
   4385   1.1  mrg 
   4386   1.1  mrg    By establishing the idea that supporting further development is "the
   4387   1.1  mrg proper thing to do" when distributing free software for a fee, we can
   4388   1.1  mrg assure a steady flow of resources into making more free software.
   4389   1.1  mrg 
   4390   1.1  mrg      Copyright (C) 1994 Free Software Foundation, Inc.
   4391   1.1  mrg      Verbatim copying and redistribution of this section is permitted
   4392   1.1  mrg      without royalty; alteration is not permitted.
   4393   1.1  mrg 
   4394   1.1  mrg 
   4395   1.3  mrg File: libgomp.info,  Node: Library Index,  Prev: Funding,  Up: Top
   4396   1.1  mrg 
   4397   1.3  mrg Library Index
   4398   1.3  mrg *************
   4399   1.1  mrg 
   4400   1.1  mrg [index]
   4401   1.1  mrg * Menu:
   4402   1.1  mrg 
   4403  1.11  mrg * Environment Variable:                  OMP_CANCELLATION.      (line 6)
   4404  1.11  mrg * Environment Variable <1>:              OMP_DISPLAY_ENV.       (line 6)
   4405  1.11  mrg * Environment Variable <2>:              OMP_DEFAULT_DEVICE.    (line 6)
   4406  1.11  mrg * Environment Variable <3>:              OMP_DYNAMIC.           (line 6)
   4407  1.11  mrg * Environment Variable <4>:              OMP_MAX_ACTIVE_LEVELS. (line 6)
   4408  1.11  mrg * Environment Variable <5>:              OMP_MAX_TASK_PRIORITY. (line 6)
   4409  1.11  mrg * Environment Variable <6>:              OMP_NESTED.            (line 6)
   4410  1.11  mrg * Environment Variable <7>:              OMP_NUM_THREADS.       (line 6)
   4411  1.11  mrg * Environment Variable <8>:              OMP_PROC_BIND.         (line 6)
   4412  1.11  mrg * Environment Variable <9>:              OMP_PLACES.            (line 6)
   4413  1.11  mrg * Environment Variable <10>:             OMP_STACKSIZE.         (line 6)
   4414  1.11  mrg * Environment Variable <11>:             OMP_SCHEDULE.          (line 6)
   4415  1.11  mrg * Environment Variable <12>:             OMP_THREAD_LIMIT.      (line 6)
   4416  1.11  mrg * Environment Variable <13>:             OMP_WAIT_POLICY.       (line 6)
   4417  1.11  mrg * Environment Variable <14>:             GOMP_CPU_AFFINITY.     (line 6)
   4418  1.11  mrg * Environment Variable <15>:             GOMP_DEBUG.            (line 6)
   4419  1.11  mrg * Environment Variable <16>:             GOMP_STACKSIZE.        (line 6)
   4420  1.11  mrg * Environment Variable <17>:             GOMP_SPINCOUNT.        (line 6)
   4421  1.11  mrg * Environment Variable <18>:             GOMP_RTEMS_THREAD_POOLS.
   4422  1.10  mrg                                                                 (line 6)
   4423  1.10  mrg * FDL, GNU Free Documentation License:   GNU Free Documentation License.
   4424   1.7  mrg                                                                 (line 6)
   4425  1.11  mrg * Implementation specific setting:       OMP_NESTED.            (line 6)
   4426  1.11  mrg * Implementation specific setting <1>:   OMP_NUM_THREADS.       (line 6)
   4427  1.11  mrg * Implementation specific setting <2>:   OMP_SCHEDULE.          (line 6)
   4428  1.11  mrg * Implementation specific setting <3>:   GOMP_STACKSIZE.        (line 6)
   4429  1.11  mrg * Implementation specific setting <4>:   GOMP_SPINCOUNT.        (line 6)
   4430  1.11  mrg * Implementation specific setting <5>:   GOMP_RTEMS_THREAD_POOLS.
   4431   1.1  mrg                                                                 (line 6)
   4432   1.1  mrg * Introduction:                          Top.                   (line 6)
   4433   1.1  mrg 
   4434   1.1  mrg 
   4435   1.1  mrg 
   4436   1.1  mrg Tag Table:
   4437  1.11  mrg Node: Top2083
   4438  1.11  mrg Node: Enabling OpenMP4489
   4439  1.11  mrg Node: Runtime Library Routines5276
   4440  1.11  mrg Node: omp_get_active_level8339
   4441  1.11  mrg Node: omp_get_ancestor_thread_num9039
   4442  1.11  mrg Node: omp_get_cancellation9969
   4443  1.11  mrg Node: omp_get_default_device10783
   4444  1.11  mrg Node: omp_get_dynamic11459
   4445  1.11  mrg Node: omp_get_level12334
   4446  1.11  mrg Node: omp_get_max_active_levels12954
   4447  1.11  mrg Node: omp_get_max_task_priority13659
   4448  1.11  mrg Node: omp_get_max_threads14279
   4449  1.11  mrg Node: omp_get_nested15036
   4450  1.11  mrg Node: omp_get_num_devices15950
   4451  1.11  mrg Node: omp_get_num_procs16471
   4452  1.11  mrg Node: omp_get_num_teams17010
   4453  1.11  mrg Node: omp_get_num_threads17526
   4454  1.11  mrg Node: omp_get_proc_bind18615
   4455  1.11  mrg Node: omp_get_schedule19536
   4456  1.11  mrg Node: omp_get_team_num20490
   4457  1.11  mrg Node: omp_get_team_size20989
   4458  1.11  mrg Node: omp_get_thread_limit21949
   4459  1.11  mrg Node: omp_get_thread_num22568
   4460  1.11  mrg Node: omp_in_parallel23439
   4461  1.11  mrg Node: omp_in_final24088
   4462  1.11  mrg Node: omp_is_initial_device24762
   4463  1.11  mrg Node: omp_set_default_device25455
   4464  1.11  mrg Node: omp_set_dynamic26246
   4465  1.11  mrg Node: omp_set_max_active_levels27132
   4466  1.11  mrg Node: omp_set_nested27909
   4467  1.11  mrg Node: omp_set_num_threads28801
   4468  1.11  mrg Node: omp_set_schedule29669
   4469  1.11  mrg Node: omp_init_lock30750
   4470  1.11  mrg Node: omp_set_lock31403
   4471  1.11  mrg Node: omp_test_lock32258
   4472  1.11  mrg Node: omp_unset_lock33234
   4473  1.11  mrg Node: omp_destroy_lock34165
   4474  1.11  mrg Node: omp_init_nest_lock34842
   4475  1.11  mrg Node: omp_set_nest_lock35577
   4476  1.11  mrg Node: omp_test_nest_lock36492
   4477  1.11  mrg Node: omp_unset_nest_lock37519
   4478  1.11  mrg Node: omp_destroy_nest_lock38534
   4479  1.11  mrg Node: omp_get_wtick39285
   4480  1.11  mrg Node: omp_get_wtime39877
   4481  1.11  mrg Node: Environment Variables40653
   4482  1.11  mrg Node: OMP_CANCELLATION42208
   4483  1.11  mrg Node: OMP_DISPLAY_ENV42741
   4484  1.11  mrg Node: OMP_DEFAULT_DEVICE43444
   4485  1.11  mrg Node: OMP_DYNAMIC44224
   4486  1.11  mrg Node: OMP_MAX_ACTIVE_LEVELS44820
   4487  1.11  mrg Node: OMP_MAX_TASK_PRIORITY45470
   4488  1.11  mrg Node: OMP_NESTED46128
   4489  1.11  mrg Node: OMP_NUM_THREADS46733
   4490  1.11  mrg Node: OMP_PROC_BIND47421
   4491  1.11  mrg Node: OMP_PLACES48613
   4492  1.11  mrg Node: OMP_STACKSIZE50790
   4493  1.11  mrg Node: OMP_SCHEDULE51614
   4494  1.11  mrg Node: OMP_THREAD_LIMIT52312
   4495  1.11  mrg Node: OMP_WAIT_POLICY52912
   4496  1.11  mrg Node: GOMP_CPU_AFFINITY53604
   4497  1.11  mrg Node: GOMP_DEBUG55334
   4498  1.11  mrg Node: GOMP_STACKSIZE55841
   4499  1.11  mrg Node: GOMP_SPINCOUNT56670
   4500  1.11  mrg Node: GOMP_RTEMS_THREAD_POOLS57874
   4501  1.11  mrg Node: Enabling OpenACC60052
   4502  1.11  mrg Node: OpenACC Runtime Library Routines61045
   4503  1.11  mrg Node: acc_get_num_devices64839
   4504  1.11  mrg Node: acc_set_device_type65565
   4505  1.11  mrg Node: acc_get_device_type66329
   4506  1.11  mrg Node: acc_set_device_num67044
   4507  1.11  mrg Node: acc_get_device_num67828
   4508  1.11  mrg Node: acc_async_test68625
   4509  1.11  mrg Node: acc_async_test_all69617
   4510  1.11  mrg Node: acc_wait70517
   4511  1.11  mrg Node: acc_wait_all71379
   4512  1.11  mrg Node: acc_wait_all_async72138
   4513  1.11  mrg Node: acc_wait_async72890
   4514  1.11  mrg Node: acc_init73597
   4515  1.11  mrg Node: acc_shutdown74243
   4516  1.11  mrg Node: acc_on_device74911
   4517  1.11  mrg Node: acc_malloc75915
   4518  1.11  mrg Node: acc_free76414
   4519  1.11  mrg Node: acc_copyin76841
   4520  1.11  mrg Node: acc_present_or_copyin77951
   4521  1.11  mrg Node: acc_create79577
   4522  1.11  mrg Node: acc_present_or_create80732
   4523  1.11  mrg Node: acc_copyout82366
   4524  1.11  mrg Node: acc_delete83389
   4525  1.11  mrg Node: acc_update_device84364
   4526  1.11  mrg Node: acc_update_self85476
   4527  1.11  mrg Node: acc_map_data86580
   4528  1.11  mrg Node: acc_unmap_data87265
   4529  1.11  mrg Node: acc_deviceptr87786
   4530  1.11  mrg Node: acc_hostptr88356
   4531  1.11  mrg Node: acc_is_present88920
   4532  1.11  mrg Node: acc_memcpy_to_device90447
   4533  1.11  mrg Node: acc_memcpy_from_device91110
   4534  1.11  mrg Node: acc_get_current_cuda_device91794
   4535  1.11  mrg Node: acc_get_current_cuda_context92391
   4536  1.11  mrg Node: acc_get_cuda_stream92985
   4537  1.11  mrg Node: acc_set_cuda_stream93541
   4538  1.11  mrg Node: OpenACC Environment Variables94073
   4539  1.11  mrg Node: ACC_DEVICE_TYPE94532
   4540  1.11  mrg Node: ACC_DEVICE_NUM94768
   4541  1.11  mrg Node: GCC_ACC_NOTIFY95025
   4542  1.11  mrg Node: CUDA Streams Usage95248
   4543  1.11  mrg Ref: CUDA Streams Usage-Footnote-197148
   4544  1.11  mrg Node: OpenACC Library Interoperability97257
   4545  1.11  mrg Ref: OpenACC Library Interoperability-Footnote-1103613
   4546  1.11  mrg Ref: OpenACC Library Interoperability-Footnote-2103865
   4547  1.11  mrg Node: The libgomp ABI104073
   4548  1.11  mrg Node: Implementing MASTER construct104929
   4549  1.11  mrg Node: Implementing CRITICAL construct105343
   4550  1.11  mrg Node: Implementing ATOMIC construct106082
   4551  1.11  mrg Node: Implementing FLUSH construct106563
   4552  1.11  mrg Node: Implementing BARRIER construct106834
   4553  1.11  mrg Node: Implementing THREADPRIVATE construct107103
   4554  1.11  mrg Node: Implementing PRIVATE clause107756
   4555  1.11  mrg Node: Implementing FIRSTPRIVATE LASTPRIVATE COPYIN and COPYPRIVATE clauses108337
   4556  1.11  mrg Node: Implementing REDUCTION clause109661
   4557  1.11  mrg Node: Implementing PARALLEL construct110218
   4558  1.11  mrg Node: Implementing FOR construct111475
   4559  1.11  mrg Node: Implementing ORDERED construct113473
   4560  1.11  mrg Node: Implementing SECTIONS construct113779
   4561  1.11  mrg Node: Implementing SINGLE construct114545
   4562  1.11  mrg Node: Implementing OpenACC's PARALLEL construct115257
   4563  1.11  mrg Node: Reporting Bugs115515
   4564  1.11  mrg Node: Copying115877
   4565  1.11  mrg Node: GNU Free Documentation License153424
   4566  1.11  mrg Node: Funding178547
   4567  1.11  mrg Node: Library Index181073
   4568   1.1  mrg 
   4569   1.1  mrg End Tag Table
   4570